Intracellular- and extracellular-derived Ca²⁺ influence phospholipase A₂-mediated fatty acid release from brain phospholipids

Summary

Docosahexaenoic acid (DHA) and arachidonic acid (AA) are found in high concentrations in brain cell membranes and are important for brain function and structure. Studies suggest that AA and DHA are hydrolyzed selectively from the sn-2 position of synaptic membrane phospholipids by Ca²⁺-dependent cytosolic phospholipase A₂ (cPLA₂) and Ca²⁺-independent phospholipase A₂ (iPLA₂), respectively, resulting in increased levels of the unesterified fatty acids and lysophospholipids. Cell studies also suggest that AA and DHA release depend on increased concentrations of Ca²⁺, even though iPLA₂ has been thought to be Ca²⁺-independent. The source of Ca²⁺ for activation of cPLA₂ is largely extracellular, whereas Ca²⁺ released from the endoplasmic reticulum can activate iPLA₂ by a number of mechanisms. This review focuses on the role of Ca²⁺ in modulating cPLA₂ and iPLA₂ activities in different conditions. Furthermore, a model is suggested in which neurotransmitters regulate the activity of these enzymes and thus the balanced and localized release of AA and DHA from phospholipid in brain, depending on the primary source of the Ca²⁺ signal.

1. Overview of the phospholipase A₂ families

Although PLA₂ enzymes have been known for over 100 years, only within the last 15 years has their importance for neurochemical processes been widely recognized [1]. These enzymes catalyze the hydrolysis of esterified fatty acids at the stereospecifically numbered (sn)-2 position of membrane phospholipids to release unesterified fatty acids and form lysophospholipids [2–9]. The fatty acids and their metabolic products are important for normal brain function and can alter neuropathological processes [1, 10–15].

PLA₂s are classified according to their sequence homology and Ca²⁺ dependence. They belong to three main families having different catalytic sequences, and these sequences are conserved among PLA₂s in the same family. The three most described families are: low molecular mass secretory PLA₂ (sPLA₂), higher molecular mass Ca²⁺-dependent cytosolic PLA₂ (cPLA₂), and “Ca²⁺-independent” PLA₂ (iPLA₂) [24](See Table 1). Other less studied PLA₂ families have been described, such as platelet-activating factor-acetyl hydrolase (PAF-AH) [8, 21] and lysosomal PLA₂ [22, 23].

Table 1.

PLA₂ enzymes and their subcellular localization, and Ca²⁺-dependence of their activity [1, 8, 16–23]

Enzyme	sPLA2	cPLA2	iPLA2	Lysosomal- PLA2	PAF-AH
			No catalytic effect.	No effect	No effect
Effect of Ca2+	Stimulated (mM)	Translocation	Activated by Ca2+ release from ER*
Expression profile	Ubiquitous	Ubiquitous	Heart, brain, placenta, islets, testes, skeletal muscle	Heart, lung, liver, kidney, spleen, brain, thymus, inflammatory cells (macrophages)	Ubiquitous
Fatty acid preference	None	AA	DHA	oleate and linoleate are preferred relative to arachidonate	DHA and AA

^*This effect may be achieved by Ca²⁺ release itself and not by Ca²⁺ from the ER binding the enzyme

The sPLA₂ family is subdivided into groups: sPLA₂-IB, sPLA₂-IIA, sPLA₂-IIC, sPLA₂-IID, sPLA₂-IIE, sPLA₂-IIF, sPLA₂-III, sPLA₂-V, sPLA2-X, and sPLA₂-XII [1]. The genes for human sPLA₂–IB map to chromosome 12q23–24, those for group II subfamily sPLA₂s (IIA, IIC, IID, IIE, IIF) and for sPLA₂-V map to chromosome 1p34–36, and the gene for sPLA₂-X maps to chromosome 16p12–13 [1]. sPLA₂s are Ca²⁺-dependent for catalytic activity (generally requiring millimolar Ca²⁺ concentrations) [1, 25, 26], consistent with their extracellular activity and millimolar Ca²⁺ concentrations in the extracellular space [27]. Evidence also points to an intracellular functional role sPLA₂ in caveolae-containing compartments around the nucleus of immune cells and in the nucleus of astrocytes and neurons [28–30].

The cPLA₂ family is subdivided into cPLA₂α, cPLA₂β, cPLA₂γ, cPLA₂δ, cPLA₂ɛ and cPLA₂ζ, which also are referred to as groups IVA, IVB, IVC, IVD, IVE and IVF, respectively. They are coded by gene on different human chromosomes, chromosome 1 (cPLA₂α), 15 (cPLA₂β, cPLA₂δ, cPLA2ɛ, cPLA2ζ) and chromosome 19 (cPLA₂γ), which are not the same as the chromosomes that code for the sPLA₂ family [31–38]. cPLA₂ becomes activated after translocating to the plasma membrane from the cytosol, and is constitutively membrane-bound [1, 38]. Although Ca²⁺ is not necessary for cPLA₂ catalytic activity, nanomolar Ca²⁺ concentrations are needed for its binding to the membrane, where other enzymes that regulate the arachidonic acid (AA) cascade, including cyclooxygenases (COXs) and lipoxygenases, also are located [16].

The iPLA₂s have extremely different structures compared to the cPLA₂ and sPLA₂ families [39]. They occur in numerous splicing variants, only some of which are functional. iPLA₂s are classified as group VI and are subdivided into VIA and VIB, or into iPLA₂β and iPLA₂γ, respectively. Each iPLA₂ is coded by a different chromosome, iPLA₂β is coded by chromosome 23q13 [40] and iPLA₂γ is coded by chromosome 7q31 [1, 41]. iPLA₂β is unique among PLA₂s, as it has multiple ankyrin repeats, which are common motifs found in more than 400 proteins, including cell cycle regulators and transcription factors. Ankyrin repeats have been implicated in protein–protein interactions and can influence many physiological functions [39]. A site near the C-terminus of some iPLA₂β variants can bind to calmodulin with the help of Ca²⁺ [1, 42–44].

The PAF-AHs are classified as VIIA (lipoprotein-associated- PLA₂), VIIB (PAF-AH II) VIIIA (PAF-Ib) and VIIIB (PAF-Ib or VIIIB forms a heterodimer with VIIIA) [5]. PAF-AH was first named for its ability to cleave the acetyl group from the sn-2 position of platelet activating factor (PAF); however, this enzyme can cleave oxidized lipids in the sn-2 position up to 9 carbons long, not just PAF. This enzyme also has been shown to access substrate in the aqueous phase, unlike the other PLA₂s studied [21], in a Ca²⁺-independent fashion [6]. Type I PAF-AH is a complex of two catalytic subunits, alpha1 and alpha2, a regulatory beta subunit, and a gamma subunit [45]. Its structure is very similar to the structure of G proteins [46], whereas type II PAF-AH is a single polypeptide [47]. Type II PAF-AH is myristoylated at the N-terminus and distributed in both the cytosol and membranes. Plasma PAF-AH has a closed structure when compared to II PAF-AH, and exists in association with plasma lipoproteins [21].

Lysosomal or group XV PLA₂ is a calcium-independent PLA₂ enzyme of 25 kDa [23], which has a distinguishing characteristic of having its maximal activity in an acidic medium, and accordingly, it also is called acidic iPLA₂ [23, 48].

This review will focus on the roles of the cPLA₂ and iPLA₂ in mammalian brain. We propose a model for receptor-mediated regulation of PLA₂ enzyme activities by Ca²⁺ signals originating from intracellular and extracellular sources, leading to the controlled, balanced and localized release of specific fatty acids as signaling molecules.

2. cPLA₂ and iPLA₂ in the brain

cPLA₂ is found in mammalian brain in three different forms: cPLA₂α (85-kDa), cPLA₂β (114-kDa), and cPLA₂γ (61-kDa). cPLA₂β is found mainly in the cerebellum, whereas cPLA₂α and cPLA₂γ are uniformly distributed in rat brain [16, 31, 33]. cPLA₂α is found both in astrocytes and neurons [16]. In neurons of rat brain, cPLA₂ has been localized at post-synaptic sites, suggesting that it is important in neurotransmission (see table 2) [17].

Table 2.

cPLA₂ and iPLA₂ enzymes found in brain [1, 17, 18, 38, 49–51]

	cPLA2	iPLA2
Localization	Cytosol	Cytosol, ER 80-kDa (iPLA2β) rat
Mol. Mass	85-kDa (cPLA2α) 114-kDa (cPLA2β) 61 -kDa (cPLA2γ)	85-kDa (subtype unknown) monkey 110-kDa (subtype unknown) bovine 30-kDa (subtype unknown) bovine

iPLA₂ is believed to account for more than 70% of brain PLA₂ activity [25]. iPLA₂β (80-kDa) was purified from rat brain [49] and found in all brain regions, with the highest activity in striatum, hypothalamus, and hippocampus [16, 49]. Similarly, two different iPLA₂s (110-kDa and 39-kDa) were purified from bovine brain [51, 52]. iPLA₂ (85-kDa) also was identified in monkey brain, in the neocortex, amygdala, hippocampus, caudate nucleus, putamen, and nucleus accumbens, whereas in the thalamus, hypothalamus and globus pallidus it was lightly labeled (see table 2). On the other hand, the midbrain, vestibular, trigeminal and inferior olivary nuclei, and the cerebellar cortex were densely labeled [18]. Individual iPLA₂ subtypes have not been investigated in the bovine or monkey brain. The enzyme is present in astrocytes [53] and in neurons, where its localization in dendrites and axon terminals suggests that it plays a role in neuronal signaling [18].

3. The cPLA₂ product: Arachidonic acid and brain function

Arachidonic acid (AA, 20:4n-6), an n-6 polyunsaturated fatty acid (PUFA), comprises approximately 5–15% of total fatty acids in most tissue phospholipids [54]. It can be synthesized in the liver from dietary derived linoleic acid, or obtained directly from dietary sources [55]. AA is preferentially released from phospholipid by cPLA₂ [1, 4, 56]. It and its metabolites (prostaglandins, leukotrienes, thromboxanes, lipoxins) in brain influence synaptic signaling [57, 58], neuronal firing [59], neurotransmitter release [60], activation of intracellular receptors [61], hypothalamic-pituitary function [62], nociception [63], gene expression [11], cerebral blood flow [64], circadian rhythm [65], and appetite [66], among other targets. Altered AA metabolism has been implicated in neuronal death [67] and in a number of neurological, neurodegenerative, and psychiatric disorders, including epilepsy [68], ischemia, stroke [69], HIV-associated dementia [13], amyotrophic lateral sclerosis [70], Alzheimer disease [71], Parkinson disease [72], schizophrenia [73], bipolar disorder [74–76] and depression [77, 78].

4. The iPLA₂ product: Docosahexaenoic acid and brain function

Docosahexaenoic acid (DHA, 22:6n-3) is the most abundant n-3 PUFA in brain and is necessary for normal brain function [54]. Esterified DHA within membrane phospholipids is hydrolyzed preferentially by iPLA₂ [79].

Humans can obtain DHA by eating fish or fish products, or can synthesize it in the liver from circulating α-linolenic acid, through serial steps of desaturation, elongation and oxidation [55, 80]42]. Dietary restriction of n-3 PUFAs in animal models decreased brain DHA content while increasing concentrations of the elongation product of AA, docosapentaenoic acid (22:5n-6) [80–82]. Rats deprived of dietary n-3 PUFAs for 15 weeks had increased brain mRNA, protein and activity levels of cPLA₂, sPLA₂ and COX-2, but decreased expression of iPLA₂ and COX-1 [83].

DHA has been found important for membrane function and fluidity, photoreceptor function [54], memory [84, 85], problem solving [86], and developmental visual and sensory functions [54]. Some studies but not others indicate that dietary DHA supplementation is beneficial in a number of brain disorders [87, 88]. In rats, dietary n-3 PUFA deprivation for 15 weeks increased aggression and depression scores on behavioral tests [89]. DHA, like AA, can be metabolized by COX and lipoxygenase enzymes and generate active compounds [90, 91]. DHA products, in contrast to AA products, seem to have a beneficial effect in inflammatory and neurodegenerative conditions. Resolvins of the D series and docosatrienes, which are bioactive DHA products that possess potent anti-inflammatory, immunoregulatory [92, 93], and neuroprotective actions [94], are termed neuroprotectins [90].

5. Docosahexaenoic acid as a signaling molecule: competition with arachidonic acid?

AA is well recognized as a signaling molecule [10, 95–97] and there are a number of lines of evidence indicating that DHA is a signaling molecule as well [12, 79, 98–100].

1) DHA can be released from phospholipid following activation of neuroreceptors

Activation of serotonin (5-HT) 5-HT_2A receptors in astroglioma cells caused DHA release from membrane phospholipids [101]. Muscarinic receptor activation by arecoline also caused DHA release and plasma-derived DHA incorporation into rat brain synaptic membranes, consistent with its participation in neuronal signaling [102, 103].

2) DHA exerts intracellular downstream effects

DHA increased N-methyl-D-aspartate (NMDA) function in neurons. Both DHA and AA increased entry of extracellular Ca²⁺ into neurons via glutamatergic NMDA receptors, and DHA also increased the probability of NMDA channel opening [104]. In another study, AA increased while DHA inhibited glutamate-induced prostaglandin release from astrocytes; DHA inhibited the AA effect [105].

Unesterified DHA can promote Ca²⁺ release from intracellular stores in the endoplasmic reticulum (ER) [106]. Ca²⁺ released from the ER can be a continuous source of Ca²⁺ for mitochondria, can activate dehydrogenases and mitochondrial ATP synthesis and energy production, and can stimulate neuroprotective signals [107]. On the other hand, Ca²⁺ uptake into mitochondria can modulate the activity of Ca²⁺ channels in the ER [108, 109]. If the balance between mitochondria and the ER is disturbed, a Ca²⁺ overload in the mitochondria can trigger necrosis or apoptosis [110].

Release of Ca²⁺ from intracellular stores via purinergic [111] or nicotinic [112] receptors can be neuroprotective. Intracellular Ca²⁺ release induced by isoflurane [113] or fructose-1,6-biphosphate was shown to be neuroprotective [114, 115]. Ca²⁺ release from the ER, induced by AA or DHA, reduced the peak of subsequent Ca²⁺ release induced by G protein activation, showing a complex relation between these two PUFAs and intracellular Ca²⁺ [100].

Neuroprotectin D₁ is a metabolite of DHA that may account for part of the opposite effects of DHA and AA. The synthesis of neuroprotectin D₁ was increased in cells exposed to the Ca²⁺ ionophore A23187 or to interleukin-1β, and depended on PLA₂ activity [98]. Both A23187 and interleukin-1β can induce Ca²⁺ release from intracellular stores [116]. However, regulation by ER Ca²⁺ of neuroprotectin D₁ synthesis induced by A23187 or interleukin-1β has not been studied. Neuroprotectin D₁ also can inhibit β-amyloid production [117] and COX-2 expression [99]. COX-2, which appears involved in neuroinflammation and is overexpressed in the Alzheimer disease brain [118, 119], converts AA to prostaglandins that also can increase β-amyloid production [120]. COX-2 also has been implicated in multiple sclerosis, amyotrophic lateral sclerosis, Parkinson disease, and Creutzfeldt-Jakob disease [14]. DHA in these conditions may modulate COX-2 activity and AA metabolite production via COX-2.

A DHA neuroprotective effect also is supported by evidence that iPLA₂ inhibition increased neuronal death induced by reactive oxygen species [121]. The data together suggest that increased iPLA₂ activity and DHA release can be neuroprotective.

In other contexts, iPLA₂ activation may have different consequences, related to its different subtypes, their location, and substrate availability in brain. In apoptotic cells, iPLA₂γ seems necessary for mitochondrial membrane pore transition formation [122]. iPLA₂ is a substrate for caspase-3 (or other caspases), and is cleaved during the apoptotic process at the consensus Asp183. The resulting fragment, iPLA₂(184-C), possesses the entire catalytic domain and seven of eight ankyrin repeats, and is functionally more active than intact iPLA₂ in cells [123]. On the other hand, inhibition of iPLA₂β induced apoptosis in one study [124]. The role of DHA in apoptosis related to iPLA₂ inhibition is not known.

6. Ca²⁺ signaling as control mechanism of iPLA₂ and cPLA₂ activities

Preferential release of DHA and AA by iPLA₂ and cPLA₂, respectively [49, 74, 79], raises the possibility that the enzymes are controlled by separate mechanisms and serve different or even opposing functions [75]. For example, pharmacological studies with chronic lithium, carbamazepine and valproate in unanesthetized rats, when combined with in vivo kinetic fatty acid modeling and molecular biology measurements, confirmed in vitro findings that cPLA₂ is specific for AA release, whereas iPLA₂ is comparatively specific for DHA release from phospholipid in brain synaptic membranes [74, 125, 126]

cPLA₂ requires Ca²⁺ for its translocation and arachidonic acid release [36]. Therefore, Ca²⁺ entry through ligand-gated ion channels, such as NMDA receptors, can activate this enzyme [127, 128]. In neurons, NMDA receptor activation induces Ca²⁺ entry from the extracellular space and activates cPLA₂, and this activation is reversed by MK-801 (an NMDA antagonist), by removing calcium from the extracellular space, or by chronically administering lithium, valproic acid or carbamazepine, mood stabilizers effective in bipolar disorder, to rats [129, 130] [131, 132]. The cPLA₂ coupling to Ca²⁺ entry through the membrane seems to depend on its location in the cell [133]. The localization of Ca²⁺ entry appears more important than the pattern of Ca²⁺ entry. In this sense, single transients, repetitive oscillationrs or sustained plateaus of Ca²⁺ activate cPLA₂. Under these circumstances, cPLA₂ translocates to the nucleus where it promotes AA release. On the other hand, potassium-induced Ca²⁺ entry can induce depolarization and the opening of voltage-dependent Ca²⁺ channels located at sites close to neurotransmitter release. However, potassium depolarization does not activate AA release, probably because cPLA₂ is not located close to releasing sites [133]. In agreement with this evidence of Ca²⁺ dependence of neurotransmitter-induced cPLA₂ activity, 5-HT receptor activation in hippocampal neurons induced an increase in PLA₂-mediated AA release that was reversed by removing extracellular Ca²⁺ [134].

iPLA₂ has been considered to be Ca²⁺-independent, because its catalytic activity was found to be independent of Ca²⁺ when studied in an incubation medium [135]. However, in fractionated cytosol, Ca²⁺ later was shown to decrease iPLA₂β activity, probably by modifying its interaction with calmodulin [43]. Multiple contact points in the 15-kDa C-terminal portion of iPLA₂β have been identified as determinants of its Ca²⁺-dependent interaction with calmodulin [44]. Interaction between calmodulin and iPLA₂ inhibits iPLA₂ activity and is modulated by Ca²⁺ [43].

iPLA₂β is highly localized in the ER in areas around the cell nucleus [20] and in the cytosol; it is less expressed near plasma membranes and mitochondria [18]. iPLA₂β immunoreactivity has been detected around the nuclear envelope in cortical and limbic areas of the rhesus brain [18], as well in dendrites and terminals, suggesting a local roles.

As illustrated in Figure 1, Ca²⁺ can be released from intracellular ER stores in several ways. The ER has two types of membrane receptors, an inositol-1,4,5-phosphate receptor (InsP₃R), and a ryanodine receptor (RyR) that can receive information from the cytosol and release Ca²⁺ from the ER lumen. This signaling can be evoked by purinergic [136, 137], muscarinic [138], or metabotropic glutamatergic receptors [139], so as to produce receptor-mediated inositol-3-phosphate (InsP₃) release and activate the InsP₃R. Activation of NMDA receptors [140, 141], AMPA receptors [142], or even InsP₃Rs [143] can release Ca²⁺ by “Ca²⁺-induced Ca²⁺ release” (CICR) through activation of RyRs. Inhibition of the reticular ATPase pump (SERCA) by thapsigargin, DBHQ (2,5-di(tert-butyl)hydroquinone), or Br(2)-TITU (1,3-dibromo-2,4,6-tris (methyl-isothio-uronium) benzene), also can cause a sustained release of Ca²⁺ from the ER by a different mechanism. Ca²⁺ is believed to be released on a continuous basis from the ER and refilled by SERCA. Blocking SERCA activity may, as a result, increase cytosolic Ca²⁺ concentrations (see Figure 1) [144]. The Ca²⁺ released from intracellular stores by ATP, A23187, thapsigargin, or arginine vasopressin may activate iPLA₂ [42, 145], even in the presence of intracellular or extracellular Ca²⁺ chelators such as BAPTA or EGTA, which suggests that ER calcium emptying can transiently modulate the enzyme availability by releasing iPLA₂ from calmodulin binding [145].

Figure 1.

Hypothetical regulation of cPLA₂ (red) and iPLA₂ (blue) activities by activation of neuronal receptors. Receptor activation can stimulate cPLA₂ though Ca²⁺ entry from extracellular space or via Gi protein coupling. iPLA₂ can be activated by Ca²⁺ release from intracellular stores through activation of InsP₃ or Ca²⁺ -induced Ca²⁺ release from the ER. (A) cPLA₂ is not stimulated by Ca²⁺ and do not release AA. SERCA works continuously to pump Ca²⁺ into the ER. In this circumstance, calmodulin inhibits iPLA₂ in a Ca²⁺ dependent manner. (B) Neuronal receptors are stimulated and permeate Ca²⁺ or stimulate the production of IP₃. cPLA₂ is translocated to the nucleus in presence of µM concentration of Ca²⁺ and release AA. When Ca²⁺ concentration is decreased inside ER, either by an increase in ER receptors stimulation or by SERCA inhibition, probably calmodulin (green, represented as CAM) is not interacting with iPLA₂, allowing its activity. Differences in Ca²⁺ symbols size represent difference in concentrations of this ion, whereas small letters represents decreased calcium concentrations and big letters represents increased calcium concentrations.

Smooth muscle cells treated with the calmodulin inhibitor W-7 or with thapsigargin release AA from the same pool, suggesting that iPLA₂β is bound to a calmodulin, and that calmodulin acts as a sensor of reticular Ca²⁺ to regulate iPLA₂ activity [42, 44]. The PLA₂ family is formed by ankyrins, and ankyrins are related to protein-protein interactions and may modulate regulation of iPLA₂ activity by calmodulin [39]. Moreover, thapsigargin increases iPLA₂β expression in purified ER [146]. Strokin et al. [79], who demonstrated the exclusive release of DHA by iPLA₂, also showed that the ionophore A23187 increased DHA release 11-fold and AA release only 3.9 -fold compared to their respective baselines. In those studies, extracellular Ca²⁺ may have stimulated cPLA₂ directly by increasing Ca²⁺ close to the plasma membrane, and have activated iPLA₂ indirectly by releasing Ca²⁺ from intracellular stores by a CICR. ATP-induced AA release was suppressed by removing extracellular Ca²⁺ in that study, but DHA release was unchanged. Thus ATP could have induced Ca²⁺ release from intracellular stores to selectively activate iPLA₂ [79, 147].

This putative modulation of iPLA₂ activity and DHA release by ER Ca²⁺ release is further suggested by experimental studies of ischemia or oxygen and glucose deprivation (OGD). In mouse C2C12 myotubes, OGD increased 4-bromoenol lactone (BEL, an iPLA₂ inhibitor)-sensitive iPLA₂ activity that could be blocked by an siRNA against iPLA₂β. iPLA₂β protein in these cells was identified mainly at the endoplasmic reticulum, where it accumulated further during OGD, whereas the mRNA level was unchanged [148]. In agreement with these results, DHA is released during decapitation-induced ischemia [149]. Ca²⁺ entry from the extracellular space is a well-known source of intracellular Ca²⁺ during OGD. However, an important component of the Ca²⁺ increase is due to ER Ca²⁺ release, and particularly SERCA dysfunction is an important mechanism for ischemic Ca²⁺ overload [150, 151]. These results suggest that OGD may activate iPLA₂ in the ER after ER Ca²⁺ release by the physical dissociation of iPLA₂ from calmodulin.

Other evidence for the activation of iPLA₂ by Ca²⁺ derived from the ER comes from a study of the iPLA₂ effect on store operated calcium entry (SOCE). SOCE is another regulatory mechanism of Ca²⁺ homeostasis. It consists in the entry of Ca²⁺ from extracellular space that is induced by Ca²⁺ released from intracellular stores [152, 153]. In rat cerebellar astrocytes, iPLA₂ is a major regulator of SOCE. During depletion of ER Ca²⁺ stores, iPLA₂ is believed to be activated, resulting in opening channels in plasma membrane by the formation of lysophospholipids, which may affect the lipid environment of the channels or directly interact with them. In this regard, both depletion of ER Ca²⁺ stores and inhibition of calmodulin increased BEL-sensitive iPLA₂ activity in cultured cerebellar astrocytes. Furthermore, the specific antisense inhibition of iPLA₂ reduces the SOCE [153]. This study, using antisense to inhibit iPLA₂ and SOCE currents as the physiological outcome, is strong evidence that iPLA₂ is indeed the PLA₂ subtype that can be dissociated from calmodulin either by an inhibitor of calmodulin or by Ca²⁺ release from ER.

These results suggest Ca²-controlled activation of iPLA₂ in vivo and different possibilities for neurotransmitter regulation of AA and DHA release and signaling, since many neurotransmitters are able to stimulate InsP₃ formation or CICR. Given that Ca²⁺ in the ER lumen is believed to be regulated by specific proteins, changes in intracellular Ca²⁺ concentrations can transmit signals both to the ER surface and the lumen. For example, calreticulin is an ER chaperone that can detect changes in Ca²⁺ inside the ER and interact with other proteins to integrate Ca²⁺ release with different cell functions [154]. Calmodulin, and possibly other Ca²⁺ sensitive proteins, may link ER Ca²⁺ release to iPLA₂ activation. Ca²⁺ from these different sources, thus, may couple neurotransmission in different ways to iPLA₂ and cPLA₂ activities, thereby regulating DHA and AA signaling.

Agonist activation of dopaminergic D₂ [155–157], serotonergic 5-HT_2A/2C [158], NMDA[130] or cholinergic muscarinic receptors [102, 103] has been shown to increase AA release from rat brain membrane phospholipids in vivo. Muscarinic and 5-HT_2A receptor activation also increased DHA release in vivo or in vitro [101–103]. D₂ and 5-HT_2A/2C agonists can activate cPLA₂ by a G protein mechanism [134, 159], whereas 5-HT agonists and NMDA agonists can do so by increasing cell entry of extracellular Ca²⁺, since cPLA₂ is Ca²⁺-dependent [133, 134]. 5-HT_2A receptor activation also can release Ca²⁺ from InsP₃R stores [160], and muscarinic agonists can increase Ca²⁺ influx into the cell to also release Ca²⁺ from intracellular stores [161]. Thus, Ca²⁺ from the two sources can separately control the activities of both cPLA₂ and iPLA₂, thereby regulating AA and DHA release from membrane phospholipids.

7. Calcium modulation of PLA₂ activity and brain pathology

In many neurodegenerative, psychiatric diseases and in conditions having increased neuronal death, the modulation exerted by DHA in docosanoid production could reduce neuronal damage (see [99] for review). Inhibition by DHA of Ca²⁺ entry through L-type calcium channels could be a beneficial effect of DHA [162]. Although a description of how differential activation of cPLA₂ and iPLA₂ by Ca²⁺ could affect pathological process is beyond the scope of this review, here we briefly present data suggesting their roles.

Alzheimer Disease

Alzheimer disease is a progressive neurodegenerative disease that can involve changes brain PLA₂ and PUFA metabolism, with evidence for marked disturbances in brain Ca²⁺ homeostasis including excess Ca²⁺ release from the ER [163–165]. AA metabolism was shown to be increased in brains of Alzheimer disease patients using positron emission tomography, presumably in relation to neuroinflammation and activation of cPLA₂ and sPLA₂ via astrocytic cytokine receptors [166, 167] [168–170], whose immunoreactivity have both been reported to be increased in the postmortem Alzheimer brain [50, 170]. cPLA₂ can be activated by entry of extracellular Ca²⁺ into cells through ionotropic NMDA receptors [127], and NMDA binding sites were decreased in the postmortem Alzheimer disease brain [171]. cPLA₂ mediated AA release can be increased by the β-amyloid peptide that accumulates in the Alzheimer disease brain, probably via mediation of NMDA receptors [172]. Additionally, a low plasma DHA level has been associated with Alzheimer disease [173], but controlled supplementation studies have not been performed. β-amyloid can induce membrane-associated oxidative stress or form an oligomeric pore in the membrane [174], which could affect cPLA₂ activity. DHA may be beneficial in Alzheimer disease by decreasing β-amyloid release [175].

Cerebral Ischemia

Depolarization caused by energy failure in cerebral ischemia will lead to opening of voltage-gated Ca²⁺ channels in the membrane and excessive release of neurotransmitters [176], including glutamate. Glutamate activation of NMDA receptors normally is limited by magnesium blocking the channel, but depolarization releases magnesium and the receptor can then be opened to Ca²⁺ by the excessive glutamate [177]. In ischemia, a sustained depolarization and excess glutamate can maintain the NMDA channel open and load Ca²⁺ into the cell [176], thereby activating Ca²⁺-sensitive enzymes including cPLA₂ [178–181], and causing inflammatory mediator production [178, 182]. Hypoxia was shown to release both AA and DHA in rat brain, at a ratio of 3.6 to 1 [149]. In ischemia, neuroprotectin D1 and DHA were protective [183–185]. DHA can block currents of L-type of calcium channels[162], and inhibit AA induced prostaglandin production [105].

Depression

Depression has been proposed to be related to NMDA receptor overactivation [186–188], with resultant neuronal damage [189]. Thus, NMDA receptor antagonists induce an antidepressant effect in animal models [190–193] and in humans [188], and classical antidepressants can regulate NMDA receptor function [186]. Neuroinflammation likely is involved in depression [194]. Decreased plasma DHA levels were found in depressed individuals with coronary problems [195], but n-3 PUFA supplementation in depressed patients produced contradictory results [196, 197].

8. iPLA₂ and brain disease

Neurodegeneration associated with brain iron accumulation (NBIA) comprises a heterogeneous group of disorders, and includes patients with mutations in the PLA2G6 gene encoding iPLA₂β [198]. Children with PLA2G6 mutations show progressive cognitive and motor skill regression, with cerebellar ataxia and dystonia, as well as cerebellar cortical atrophy and gliosis. Mice with a targeted disruption of the iPLA2β gene show severe motor dysfunction, associated with widespread degeneration of brain axons and/or synapses, accompanied by the swollen axons and vacuoles [199].

9. Conclusions

cPLA₂ and iPLA₂ are important lipid regulatory enzymes that selectively release AA and DHA, respectively, from brain membrane phospholipids. The two PUFAs and their metabolites have many physiological effects and their relative rates and brain sites of release, controlled by the two enzymes and their concentrations in phospholipid, may help to fine-tune brain physiology and metabolism and when disrupted, lead to psychiatric and neurological disease. cPLA₂ has been shown to depend on extracellular-derived Ca²⁺ in both in vitro and in vivo studies. Although Ca²⁺ was shown not to be necessary for iPLA₂ activation in some in vitro studies, Ca²⁺ released from the intracellular stores of the ER can activate iPLA₂ in cells, which suggests that activation can depend on Ca²⁺ in vivo. Neurotransmitters coupled to InsP₃R activation or to CICR modulate iPLA₂ activity by regulating intracellular Ca²⁺ stores. Further experiments are needed to elaborate the in vivo conditions under which this occurs.

Abbreviations

AA

arachidonic acid

AMPA

α-amino-3-hydroxy-5-methylisoxazole-4-propionic acid

BEL

4-bromoenol lactone (inhibitor of iPLA₂)

DHA

docosahexaenoic acid

ER

endoplasmic reticulum

CICR

Ca²⁺-induced Ca²⁺ release

COX

cyclooxygenase

InsP₃R

inositol 1,4,5-trisphosphate receptor

LOX

lipoxygenase

NMDA

N-methyl-D-aspartate

OGD

oxygen and glucose deprivation

PAF-AH

Plasma activated factor-acetyl hydrolase PLA₂, phospholipase A₂

cPLA₂

cytosolic phospholipase A₂

iPLA₂

Ca²⁺-independent phospholipase A₂

sPLA₂

secretory phospholipase A₂

PUFA

polyunsaturated fatty acid

5-HT

5-hydroxytrypamine (serotonin)

RyR

ryondine receptor

SERCA

sarco(endo)plasmic reticulum Ca²⁺ ATPase

sn-2 position

position 2 in a 3-carbon glycerol backbone