Review

TRENDS in Neurosciences

Vol.31 No.1

Water movements in the brain: role of aquaporins
Matthew J. Tait, Samira Saadoun, B. Anthony Bell and Marios C. Papadopoulos
Academic Neurosurgery Unit, St. George’s University of London, Cranmer Terrace, Tooting, London SW17 0RE, UK

About 80% of the brain is water. This review discusses the importance of the three brain water-channel proteins (AQP1, AQP4, AQP9) in brain physiology. AQP1 is expressed in the choroid plexus and participates in forming cerebrospinal fluid. AQP4, found in astrocyte foot processes, glia limitans and ependyma, facilitates water movement into and out of the brain, accelerates astrocyte migration and alters neuronal activity. Recently, AQP4 autoantibodies were discovered in patients with neuromyelitis optica, a demyelinating disease, and are now being used to diagnose this condition. AQP9 is present in some glia and neurons, but its function is unclear. Finally, we discuss how the discovery of AQP activators and inhibitors will be the next major step in this field. Water channels: general properties Although 80% of the brain is water, relatively little is known about brain water physiology. We now know about a family of water-channel proteins, called aquaporins (AQPs), which increase plasma membrane osmotic permeability. The existence of AQPs was suspected long before their identification from experiments showing that red blood cell membranes are more permeable to water than expected from water diffusion through a lipid bilayer [1]. In 1988, Peter Agre discovered the first water channel, termed AQP1 [2], and was awarded the Nobel Prize in chemistry in 2003. He subsequently showed that frog oocytes expressing AQP1 in their plasma membrane were far more susceptible to osmotic lysis than nonexpressing oocytes [3], thus proving that AQP1 transports water. To date, at least 13 AQPs have been found in mammals and more than 300 in lower organisms. The AQPs are proteins that assemble in cell membranes as tetramers [4,5]. Each monomer is 30 kDa and has six membrane-spanning domains surrounding a water pore that can transport water in both directions (Figure 1). AQPs selectively pass water, except for the aquaglyceroporins (AQP3, AQP7, AQP9), which also pass glycerol and some polar molecules. Most of our knowledge of the roles of AQPs in different tissues comes from experiments with AQP-null mice (reviewed in Ref. [4]). In general, AQPs increase water permeability across epithelia, thus allowing fast water flow to accompany active salt transport. This principle is illustrated by AQP5 deletion in mice, which causes the salivary and airway submucosal glands to
Corresponding author: Papadopoulos, M.C. (mpapadop@sgul.ac.uk). Available online 4 December 2007.
www.sciencedirect.com

secrete a low volume of relatively hypertonic fluid [6,7]. AQPs also facilitate water flow in response to passive osmotic gradients. For example, AQPs increase osmotic reabsorption of water from renal collecting ducts and, as a result, AQP-deficient mice (AQPs 2–4) have impaired ability to concentrate urine [8]. Several aquaglyceroporin functions have been described, related to their glycerol-transporting ability. In humans and rodents, AQP3 is found in skin keratinocytes. AQP3-null mice have a dry, scaly skin because of reduced glycerol content in the stratum corneum and epidermis [9,10]. AQP7 is expressed in fat cells, and mice lacking AQP7 carry more fat than wild-type mice [11]. Because cell membrane glycerol permeability of AQP7-deficient fat cells is reduced, glycerol accumulates intracellularly and stimulates triglyceride synthesis. AQP expression in brain Aquaporin-4 AQP4, the principal AQP in mammalian brain, was first cloned from rat lung [12] and was found in electrically excitable tissues including brain, spinal cord, retina, inner ear and skeletal muscle [12–18]. A general paradigm is that AQP4 is not expressed in excitable cells, but is found in supporting cells (astrocytes and ependyma in the central nervous system [CNS]; Muller glia in the retina; Hensen’s, ¨ Claudius and inner sulcus cells in the ear). Brain AQP4 is strongly expressed at the borders between brain parenchyma and major fluid compartments including astrocyte foot processes (brain–blood), glia limitans (brain–subarachnoid cerebrospinal fluid [CSF]), as well as ependymal cells and

Glossary
a-syntrophin: Intracellular protein that may bind AQP4 and Kir4.1. AQP1: Found in choroid plexus epithelium. AQP4: Found in astrocyte foot processes, glia limitans, ependyma and subependymal astrocytes. AQP9: Found in some glia and neurons. Astrocytoma: Infiltrating astrocyte tumour, histologically graded I–IV. Glia limitans externa: Astrocyte processes bordering brain and subarachnoid CSF. Glia limitans interna: Astrocyte processes and ependyma bordering brain and ventricular CSF. Kir4.1: Inward rectifying K+ channel that colocalizes with AQP4. Lamellipodium: A sheet-like cytoplasmic extension produced by migrating cells. M1, M23: AQP4 isoforms. Neuromyelitis optica: Autoimmune disease characterized by optic neuritis and myelitis. Oedema: Excess water accumulation in tissue (see Box 1). Orthogonal array: Square lattice arrangement of AQP4 tetramers in cell membrane.

0166-2236/$ – see front matter ß 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.tins.2007.11.003

38

Review

TRENDS in Neurosciences

Vol.31 No.1

Figure 1. Schematic showing AQP1 structure. (a) AQP1 monomer consists of six a-helical domains (H1–H6) around a water pore, two conserved NPA motifs (Asn-Pro-Ala) that allow water but not small solutes to pass across the pore, and intracellular N and C termini. (b) Tetrameric arrangement of AQP1 in membrane. Each monomer has a water pore.

subependymal astrocytes (brain–ventricular CSF) [14,15]. This pattern of distribution suggests that AQP4 controls water flow into and out of the brain. Although AQP4 expression is polarized in astrocyte foot processes adjacent to endothelial cells, in the absence of endothelia (e.g. cultured astrocytes [19], malignant astrocytes [20,21], glial limitans and circumventricular organs [14,15]), AQP4 redistributes throughout the astrocyte cell membrane. This suggests that endothelial cells signal astrocytes to polarize AQP4 expression in the cell membrane. Two AQP4 splice variants have been cloned from brain, M1 and M23, which form square arrays of tetramers within the cell membrane [22,23]. According to a recent study, the relative amounts of M1 versus M23 isoforms determine the size of these arrays, which can contain up to hundreds of tetramers [23]. It is not known why AQP4 forms orthogonal arrays. We propose that orthogonally interconnected tetramers allow more efficient anchoring of AQP4 to intracellular proteins, such as a-syntrophin [24], compared with noninteracting AQP4 tetramers. For example, to restrict AQP4 expression to astrocyte endfeet, each noninteracting AQP4 tetramer would need its own intracellular anchor, compared with one anchor per array. There are several, often inconsistent, reports of altered AQP4 expression in different diseases. Such inconsistencies probably arise because of different AQP detection methods (mRNA versus protein), antibody specificities, tissue processing, species-specific responses and insult severities. In general, AQP4 expression is upregulated in astrocytes associated with brain oedema. For example, AQP4 overexpression in human astrocytomas correlates with the presence of brain oedema on magnetic resonance scans [20,21]. Correlations between AQP4 expression and brain oedema have been reported after brain ischaemia [25,26] and traumatic brain injury [25,27] in rodents. Such studies reinforce the notion that AQP4 plays an important role in the formation and/or absorption of brain oedema fluid. AQP4 expression also becomes upregulated in reactive astrocytes and reactive microglia [19,28], suggesting a possible role in glial scar formation. Overall, regulation studies provide only indirect evidence about the functions of AQP4 in the brain.
www.sciencedirect.com

A recent intriguing discovery is the detection of AQP4 autoantibodies in the sera of human patients with neuromyelitis optica (NMO; Devic’s syndrome), a demyelinating condition [29]. NMO serum immunostained mouse brain in a pattern corresponding to AQP4 distribution, with no immunoreactivity in AQP4-null mice. NMO serum also immunostained AQP4-expressing, but not AQP4-lacking, cultured cells. This exciting finding has led to the development of an objective diagnostic test for NMO. Some publications have suggested that AQP4 autoantibodies cause NMO, probably by inhibiting AQP4 [30,31]. To prove causality, however, it is essential to show that administering the autoantibody (e.g. to mice) causes NMO, which has not been demonstrated. The idea that AQP4 autoantibodies cause NMO raises several questions (Box 1). Despite reservations, the discovery of AQP4 autoantibodies in NMO is an interesting observation and suggests that autoantibodies against other AQPs may also play a role in other diseases. Aquaporin-1 AQP1 is expressed in the apical membrane of the choroid plexus, and plays a role in forming CSF [32]. AQP1 is upregulated in choroid plexus tumours [33], which are

Box 1. Does AQP4 autoantibody cause NMO? Outstanding questions 
Although AQP4 is expressed in kidneys, lung, inner ear, muscle and intestine, these organs are unaffected in NMO, even though they are more accessible to AQP4 autoantibodies than the CNS. Even if we assume that AQP4 function is unimportant in these organs, why does AQP4 autoantibody binding and downstream destructive events (such as complement activation) not damage these tissues?  Although AQP4 is ubiquitously expressed throughout the CNS, why does NMO primarily cause optic and spinal cord lesions?  Peripherally formed AQP4 autoantibodies cannot enter brain. Is there a preceding event that opens the BBB?  Is AQP4 channel inhibition by AQP4 autoantibody necessary for NMO to develop? AQP4 deletion does not cause NMO-like phenotype in mice.  Does serum from patients with NMO cause NMO-type disease when given to wild-type, but not to AQP4-null, mice?

Review

TRENDS in Neurosciences

Vol.31 No.1

39

associated with increased CSF production, again supporting a role for AQP1 in CSF secretion. Interestingly, AQP1 is not found in normal brain capillary endothelium, although highly expressed in peripheral endothelial cells [34,35]. Cerebral capillary endothelial cells cultured in the absence of astrocytes [34] and capillary endothelial cells within brain tumours (which are not surrounded by astrocyte endfeet) express AQP1 [36]. These observations suggest that astrocyte endfeet may signal adjacent endothelial cells to switch off endothelial AQP1 expression, although the precise signalling remains unknown. AQP1 is also found in smalldiameter sensory neurons in dorsal root, trigeminal and nodose ganglia, but studies of a possible role for AQP1 in pain sensation have yielded conflicting results [37,38]. Aquaporin-9 AQP9 immunoreactivity in rat brain has been observed in some white-matter astrocytes, Bergmann glia and a subpopulation of glia called tanycytes (reviewed in Ref. [39]). Because AQP9 transports energy substrates (glycerol, lactate), AQP9 may play a role in controlling cerebral energy metabolism. Further support for this idea comes from reports of AQP9 immunoreactivity in glucose-sensitive catecholaminergic neurons [39] and in mitochondria [40]. In general, AQP9 protein expression becomes upregulated in several brain diseases (reviewed in Ref. [39]), for example in astrocytes bordering cerebral infarcts in mice. It has been suggested that AQP9 facilitates the clearance of lactate during reperfusion. We urge caution, however, when interpreting AQP9 protein expression studies. The comparable AQP9 immunoreactivities seen in the brains of wild-type and AQP9-null mice [41] suggest that AQP9 antibodies might crossreact with other proteins. AQP functions in brain Experiments comparing wild-type versus AQP4-null mice (or a-syntrophin-null mice, which have disrupted AQP4 expression) revealed three major roles for AQP4 in brain: AQP4 controls water movements into and out of the brain, facilitates astrocyte migration and alters neuronal activity. AQP1 plays a role in forming CSF. Because of the lack of AQP9-null mice until recently, the functions of AQP9 in brain are less well characterized. Water movements into and out of brain The normal adult human intracranial cavity (1.4 L) comprises several compartments including blood (100 mL), CSF (100 mL) and brain parenchyma intracellular (1.1 L) and interstitial (100 mL) spaces. Water flows between these compartments in response to osmotic and hydrostatic forces. In several diseases, the brain swells because excess water accumulates in the brain parenchyma. To understand the role of AQP4 in brain swelling, we need to consider the main types of brain oedema, cytotoxic, vasogenic and hydrocephalic, defined in Box 2. There is mounting evidence that AQP4 deficiency is associated with reduced cytotoxic brain oedema in mouse models, including water intoxication, focal cerebral ischaemia [42] and bacterial meningitis [43]. For example, AQP4null mice develop 35% less hemispheric brain swelling than wild-type mice 24 h after middle cerebral artery
www.sciencedirect.com

Box 2. Brain oedema
The main types of brain oedema are (i) cytotoxic, (ii) vasogenic and (iii) hydrocephalic, with each involving different routes of water entry into brain [69]. (i) Cytotoxic oedema is cell swelling, associated with a reduced ECS volume and intact BBB. Water flows from the vasculature into the intracellular brain compartment when the Na+/K+ ATPase fails or extracellular [Na+] falls. Cytotoxic oedema is produced by earlyphase cerebral ischaemia, hypoxia and hyponatraemia. (ii) In vasogenic oedema, hydrostatic forces cause extravasation of a protein-rich exudate from plasma, through leaky BBB into brain ECS, causing ECS expansion. Brain tumour and brain abscess produce vasogenic oedema. (iii) Hydrocephalic oedema arises when CSF pressure is high, causing CSF extravasation through ependyma into periventricular brain. The excess fluid is evident on computed tomogram and magnetic resonance scans as ‘periventricular lucencies.’ As the brain swells inside the noncompliant skull, intracranial pressure rises, causing brain ischaemia, herniation and eventually death.

occlusion [42]. Reduced brain swelling after cerebral ischaemia and water intoxication has also been reported in a-syntrophin-null mice, which have reduced AQP4 expression in astrocyte foot processes [44,45]. Compared with wild-type mice, in AQP4- [43] and a-syntrophin- [44] deficient mice the rate of water uptake into brain parenchyma after water intoxication is reduced ninefold [43]. Water molecules moving from the blood, through intact blood–brain barrier (BBB), into brain cross three cell membranes: luminal and basal endothelial membranes that lack AQPs, and astrocyte foot process membrane that contains AQP4. Water moves across the endothelium by simple diffusion and vesicular transport, and across the astrocyte foot process primarily through AQP4 channels. These membranes are equivalent to three resistors in series, with the astrocyte membrane resistance substantially increased in AQP4 deficiency. The routes for water entry into brain are shown in Figure 2. In contrast to its beneficial role in cytotoxic oedema, AQP4 deficiency produces more brain swelling in mouse models of vasogenic oedema, including brain tumour, infusion of normal saline into brain extracellular space (ECS) and focal cortical freeze injury (that renders the BBB leaky) [46]. Melanoma cells implanted in brains of wildtype and AQP4-null mice produce comparable tumours (45 mm3) after a week with higher intracranial pressures in AQP4-null mice (40 versus 20 cmH2O). These findings

Figure 2. Routes of water flow into the brain. (a) In cytotoxic oedema, water moves from the blood through endothelial cells and astrocyte foot process membrane (through AQP4) into brain. (b) In vasogenic oedema, water moves from blood down a hydrostatic pressure gradient through a leaky BBB into brain ECS. Water entry is not through AQP4.

40

Review

TRENDS in Neurosciences

Vol.31 No.1

suggest that in vasogenic brain oedema, excess water enters brain ECS independently of AQP4, but exits the brain primarily through AQP4 water channels into the CSF and blood (Figure 3). This is a radical proposition because it implies that interstitial water is eliminated by an AQP4dependent transmembrane route, rather than by bulk flow between cells as originally thought [47,48]. After kaolin injection into the cisterna magna, which causes obstructive hydrocephalus, AQP4-null mice developed more severe hydrocephalus than wild-type mice [49]. Reduced interstitial water clearance from brains of AQP4-null mice is consistent with the hydrodynamic theory of CSF flow [50]. According to this theory, excess interstitial fluid is eliminated primarily by a trans-capillary (AQP4-rich) route into the blood rather than by bulk flow into the CSF and arachnoid granulations. The routes of water exit from the brain are shown in Figure 3. Based on the above, we propose that AQP4 inhibitors may reduce cytotoxic brain swelling in humans, whereas AQP4 activators or upregulators may reduce vasogenic oedema and hydrocephalus. Although such compounds are not currently available, their clinical usefulness cannot be overemphasized. Brain oedema is a major cause of death and disability in various neurological conditions including stroke, brain tumour and brain injury. Current treatments are inadequate and involve administering osmotic diuretics with transient benefit, corticosteroids or performing invasive procedures such as decompressive craniectomy.

CSF formation There is direct evidence supporting a role for AQP1 in CSF secretion by the choroid plexus. A dye dilution technique showed 25% reduced CSF formation rate in AQP1-null versus wild-type mice [32]. AQP1-null mice also had reduced osmotic permeability of the choroid plexus epithelium and decreased intracranial pressure [32]. We therefore suggest that AQP1 inhibitors might be useful adjunct treatments for hydrocephalus and benign intracranial hypertension, which are associated with increased CSF formation or accumulation. Astrocyte migration An unanticipated role for AQPs in facilitating cell migration was first reported for AQP1 and vascular endothelial cells in mice implanted with melanoma tumours [51]. Subsequent studies showed that AQPs enhance cell migration independent of which AQP or cell type is involved (reviewed in Ref. [52]). In general, lack of AQP slows migration speed two- to threefold [19,51,52]. Impaired migration in cultured AQP4 knockout and knockdown astrocytes, demonstrated using monolayer scratch and transwell assays, was associated with impaired glial scarring in AQP4-null mice in vivo [19]. In migrating cultured astrocytes, AQP4 polarized to the leading cell membrane edge and was associated with increased turnover of cell membrane protrusions, suggesting a role for AQP4 at the front end of migrating cells. In a follow-up study, cortical astrocytes cultured from wild-type and AQP4-null mice

Figure 3. Routes of water flow out of the brain. (a) Excess water moves from brain, through several layers of astrocyte processes comprising the glia limitans externa and pia into subarachnoid CSF. (b) Excess water also moves from brain, through layers of subependymal astrocyte processes (glia limitans interna) and ependyma into ventricular CSF. (c) Some excess water moves from brain through astrocyte foot processes and endothelial cells into the blood. All exit routes strongly express AQP4.
www.sciencedirect.com

Review

TRENDS in Neurosciences

Vol.31 No.1

41

According to the second theory, migrating cells undergo rapid changes in cell shape and cell volume as they move through the irregularly shaped ECS between stationary cells. AQPs accelerate cell migration by facilitating the transmembrane water fluxes that mediate such cell volume changes. This theory explains why differences in cell migration between wild-type and AQP4-null astrocytes are more marked in brain than in culture and why wild-type astrocytes migrating through brain ECS have more elongated shapes than migrating AQP4-null astrocytes [19,28]. The involvement of AQPs in cell migration has exciting therapeutic implications. First, because glial scarring is a major obstacle to neuronal regeneration [53], AQP4 inhibitors may enhance axonal sprouting and synaptogenesis after CNS injury by delaying glial scarring. Second, AQP4 overexpression is a feature of astrocytomas, which are highly infiltrative tumours [20,54]. AQP upregulation has also been detected in other aggressive human tumours [55,56], and a recent study found more metastases in mice after intravenous injection with AQP1-expressing versus non-AQP-expressing tumour cells [57]. We propose that AQPs facilitate cancer spread, and thus AQP inhibitors may slow tumour growth. Neuronal activity In the 1990s, morphological studies showed colocalisation of AQP4 with the inward rectifying channel Kir4.1 in the retina, leading to the idea that water and K+ fluxes are coupled [58]. More recently, AQP4-null mice were found to have increased auditory brainstem evoked response thresholds [17] and reduced electroretinographic potentials [16]. Compared with wild-type mice, AQP4-null mice have higher seizure threshold, but longer seizures in response to the convulsant pentylenetetrazol [59] or hippocampal stimulation [60]. Similar findings have been reported in a-syntrophin-null mice (which have altered AQP4 expression) [24]. There is also evidence of altered neurotransmitter release after high K+ depolarisation in AQP4-null mouse brain [61]. Taken together, these studies suggest a major role for AQP4 in controlling neuronal activity. It has been proposed that altered K+ kinetics in brain ECS account for the altered neuronal activity in AQP4 deficiency. Several lines of evidence support this notion. In vivo, AQP4-null mice have delayed K+ clearance from brain ECS after local electrical stimulation [60] and during cortical excitation by spreading depression [62]. Hippocampal slices from a-syntrophin-null mice also have slowed K+ kinetics after evoked neuronal activity [24]. To explain the altered K+ kinetics in AQP4 deficiency, a functional association between AQP4-facilitated water movement and K+ movement through the Kir4.1 channel has been proposed. Although there is a close physical association between AQP4 and Kir4.1 in astrocytes and Muller cells [24,58], patch-clamp studies in astrocytes [63] ¨ and Muller cells [64] did not find functionally significant ¨ interactions between AQP4 and Kir4.1. An alternative mechanism (ECS expansion) has recently been proposed, to account for higher seizure threshold and prolonged seizure duration in AQP4 deficiency. According to this idea, increased ECS volume increases the buffering capacity for K+ released into the ECS during neuronal

Figure 4. Two possible mechanisms of AQP4 involvement in astrocyte migration. (a) (i) Actin depolymerisation and ion transport increase intracellular osmolality in the lamellipodium. (ii) Water enters the lamellipodium through the cell membrane, facilitated by AQP4. This increases local cytoplasmic pressure forming a cell membrane protrusion. (iii) Actin repolymerizes to stabilize the protrusion. (b) In migrating astrocytes, AQP4 facilitates transmembrane water movements, which are required to allow rapid changes in cell shape and volume to occur as the cell squeezes through the extracellular space between stationary cells.

were fluorescently tagged, implanted into mouse cortex and induced to migrate toward a stab wound [28]. Two days after implantation, AQP4-null astrocytes migrated shorter distances within brain than wild-type astrocytes. Molecular mechanisms responsible for AQP involvement in cell migration are unknown. Two theories have recently been suggested (Figure 4) [52]. According to the first theory, actin depolymerisation and ion influx at the leading edge of migrating cells increase local cytoplasm osmolality, thus promoting water influx. Water entry through the cell membrane, facilitated by AQPs, causes a local rise in hydrostatic pressure that expands the cell membrane. This is followed by actin repolymerisation, which stabilizes the membrane protrusion. This theory accounts for the increased cell membrane ruffling observed at the front end of AQP-expressing migrating astrocytes, and for the enhanced migration of astrocytes toward hypo-osmolality when placed in osmotic gradients [19].
www.sciencedirect.com

42

Review

TRENDS in Neurosciences

Vol.31 No.1

Table 1. Potential clinical uses of AQP modulators in diseases of the brain
Disease process Cytotoxic oedema Vasogenic oedema CNS injury Brain tumour Seizures CSF flow abnormalities Example Ischaemic stroke Tumour, abscess Brain/spinal cord trauma Astrocytoma Epilepsy Hydrocephalus, BIH Drug AQP4 AQP4 AQP4 AQP4 AQP1 AQP4 AQP1 inhibitor activator/upregulator inhibitor inhibitor inhibitor inhibitor inhibitor Effect Reduces brain swelling Reduces brain swelling Delays glial scarring, increases neuronal regeneration Reduces tumour infiltration Reduces tumour vascularity/growth Increases seizure threshold Reduces CSF production rate

Abbreviations: BIH, benign intracranial hypertension; CNS, central nervous system; CSF, cerebrospinal fluid.

excitation, preventing large changes in ECS [K+]. Recent experiments revealed increased diffusion of fluorescently labelled macromolecules in mouse superficial cerebral cortex [65] and deep in brain [66], consistent with an expanded ECS in AQP4 deficiency. It is unclear, however, why the ECS volume is increased in AQP4-null mice, and whether ECS expansion can quantitatively account for the differences in K+ fluxes into and out of the ECS. Brain energy metabolism Major roles for AQP9 in brain energy metabolism have recently been proposed (reviewed in Ref. [39]). Lactate transport across astrocyte cell membranes, facilitated by AQP9, is consistent with the shuttle hypothesis whereby lactate, produced by astrocytes in a neuron-activity-dependent manner, diffuses to neurons for energy consumption [67]. The discovery of AQP9 immunoreactivity in glucosesensitive neurons [39] and mitochondria [40] further supports a role for AQP9 in brain energy metabolism. However, some investigators have doubted the presence of AQPs in mitochondria, based on functional assays and theoretical arguments [68]. The recently produced AQP9-null mouse should shed light on the functions of AQP9 in the brain [41]. Conclusions During the past two decades, our understanding of brain water physiology has progressed from the discovery of AQPs to the characterisation of their physiological and pathological functions. This body of work suggests that AQP modulation may have several clinical uses, summarized in Table 1. Although AQP-modulating drugs are not currently available, studies are under way using high-throughput screening of chemical libraries to discover AQP4 modulators. Another strategy is to design AQP modulators in silico using information from the AQP crystal structure. AQP4 downregulation could also be achieved by RNA inhibition with local delivery of small inhibitory RNA. We hope that in the next decade we will witness the discovery of AQP modulators to be used for treating several brain conditions including trauma, tumour, hydrocephalus and seizures.
Acknowledgements
We thank the Neurosciences Research Foundation (M.C.P., B.A.B., M.J.T.), the Royal College of Surgeons of England (M.J.T.) and the Wellcome Trust (M.C.P., S.S.) for supporting our research.

References
1 Sidel, V.W. and Solomon, A.K. (1957) Entrance of water into human red cells under an osmotic pressure gradient. J. Gen. Physiol. 41, 243–257 2 Denker, B.M. et al. (1988) Identification, purification, and partial characterization of a novel Mr 28,000 integral membrane protein from erythrocytes and renal tubules. J. Biol. Chem. 263, 15634–15642
www.sciencedirect.com

3 Preston, G.M. et al. (1992) Appearance of water channels in Xenopus oocytes expressing red cell CHIP28 protein. Science 256, 385–387 4 Verkman, A.S. (2005) More than just water channels: unexpected cellular roles of aquaporins. J. Cell Sci. 118, 3225–3232 5 Agre, P. et al. (2002) Aquaporin water channels—from atomic structure to clinical medicine. J. Physiol. 542, 3–16 6 Song, Y. et al. (2002) Localization of aquaporin-5 in sweat glands and functional analysis using knockout mice. J. Physiol. 541, 561–568 7 Song, Y. and Verkman, A.S. (2001) Aquaporin-5 dependent fluid secretion in airway submucosal glands. J. Biol. Chem. 276, 41288–41292 8 Verkman, A.S. (2006) Roles of aquaporins in kidney revealed by transgenic mice. Semin. Nephrol. 26, 200–208 9 Hara, M. et al. (2002) Selectively reduced glycerol in skin of aquaporin3-deficient mice may account for impaired skin hydration, elasticity, and barrier recovery. J. Biol. Chem. 277, 46616–46621 10 Hara, M. and Verkman, A.S. (2003) Glycerol replacement corrects defective skin hydration, elasticity, and barrier function in aquaporin3-deficient mice. Proc. Natl. Acad. Sci. U. S. A. 100, 7360–7365 11 Hara-Chikuma, M. et al. (2005) Progressive adipocyte hypertrophy in aquaporin-7-deficient mice: adipocyte glycerol permeability as a novel regulator of fat accumulation. J. Biol. Chem. 280, 15493–15496 12 Hasegawa, H. et al. (1994) Molecular cloning of a mercurial-insensitive water channel expressed in selected water-transporting tissues. J. Biol. Chem. 269, 5497–5500 13 Yang, B. et al. (2000) Skeletal muscle function and water permeability in aquaporin-4 deficient mice. Am. J. Physiol. Cell Physiol. 278, C1108–C1115 14 Rash, J.E. et al. (1998) Direct immunogold labeling of aquaporin-4 in square arrays of astrocyte and ependymocyte plasma membranes in rat brain and spinal cord. Proc. Natl. Acad. Sci. U. S. A. 95, 11981–11986 15 Nielsen, S. et al. (1997) Specialized membrane domains for water transport in glial cells: high-resolution immunogold cytochemistry of aquaporin-4 in rat brain. J. Neurosci. 17, 171–180 16 Li, J. et al. (2002) Mildly abnormal retinal function in transgenic mice without Muller cell aquaporin-4 water channels. Invest. Ophthalmol. ¨ Vis. Sci. 43, 573–579 17 Li, J. and Verkman, A.S. (2001) Impaired hearing in mice lacking aquaporin-4 water channels. J. Biol. Chem. 276, 31233–31237 18 Oshio, K. et al. (2004) Expression of aquaporin water channels in mouse spinal cord. Neuroscience 127, 685–693 19 Saadoun, S. et al. (2005) Involvement of aquaporin-4 in astroglial cell migration and glial scar formation. J. Cell Sci. 118, 5691–5698 20 Saadoun, S. et al. (2002) Aquaporin-4 expression is increased in oedematous human brain tumours. J. Neurol. Neurosurg. Psychiatry 72, 262–265 21 Warth, A. et al. (2007) Expression pattern of the water channel aquaporin-4 in human gliomas is associated with blood-brain barrier disturbance but not with patient survival. J. Neurosci. Res. 85, 1336–1346 22 Rash, J.E. et al. (2004) Freeze-fracture and immunogold analysis of aquaporin-4 (AQP4) square arrays, with models of AQP4 lattice assembly. Neuroscience 129, 915–934 23 Furman, C.S. et al. (2003) Aquaporin-4 square array assembly: opposing actions of M1 and M23 isoforms. Proc. Natl. Acad. Sci. U. S. A. 100, 13609–13614 24 Amiry-Moghaddam, M. et al. (2003) Delayed K+ clearance associated with aquaporin-4 mislocalization: phenotypic defects in brains of a-syntrophin-null mice. Proc. Natl. Acad. Sci. U. S. A. 100, 13615–13620 25 Chen, C.H. et al. (2007) Effect of osmotherapy with hypertonic saline on regional cerebral edema following experimental stroke: a study utilizing magnetic resonance imaging. Neurocrit. Care 7, 92–100

Review

TRENDS in Neurosciences

Vol.31 No.1

43

26 Meng, S. et al. (2004) Correspondence of AQP4 expression and hypoxicischaemic brain oedema monitored by magnetic resonance imaging in the immature and juvenile rat. Eur. J. Neurosci. 19, 2261–2269 27 Sun, M.C. et al. (2003) Regulation of aquaporin-4 in a traumatic brain injury model in rats. J. Neurosurg. 98, 565–569 28 Auguste, K.I. et al. (2007) Greatly impaired migration of implanted aquaporin-4-deficient astroglial cells in mouse brain toward a site of injury. FASEB J. 21, 108–116 29 Lennon, V.A. et al. (2005) IgG marker of optic-spinal multiple sclerosis binds to the aquaporin-4 water channel. J. Exp. Med. 202, 473–477 30 Matiello, M. et al. (2007) Neuromyelitis optica. Curr. Opin. Neurol. 20, 255–260 31 Wingerchuk, D.M. (2007) Neuromyelitis optica: new findings on pathogenesis. Int. Rev. Neurobiol. 79, 665–688 32 Oshio, K. et al. (2005) Reduced cerebrospinal fluid production and intracranial pressure in mice lacking choroid plexus water channel aquaporin-1. FASEB J. 19, 76–78 33 Longatti, P. et al. (2006) Aquaporin(s) expression in choroid plexus tumours. Pediatr. Neurosurg. 42, 228–233 34 Dolman, D. et al. (2005) Induction of aquaporin 1 but not aquaporin 4 messenger RNA in rat primary brain microvessel endothelial cells in culture. J. Neurochem. 93, 825–833 35 Nielsen, S. et al. (1993) Distribution of the aquaporin CHIP in secretory and resorptive epithelia and capillary endothelia. Proc. Natl. Acad. Sci. U. S. A. 90, 7275–7279 36 Saadoun, S. et al. (2002) Increased aquaporin 1 water channel expression in human brain tumours. Br. J. Cancer 87, 621–623 37 Oshio, K. et al. (2006) Impaired pain sensation in mice lacking aquaporin-1 water channels. Biochem. Biophys. Res. Commun. 341, 1022–1028 38 Shields, S.D. et al. (2007) Anatomical and functional analysis of aquaporin 1, a water channel in primary afferent neurons. Pain 131, 8–20 39 Badaut, J. and Regli, L. (2004) Distribution and possible roles of aquaporin 9 in the brain. Neuroscience 129, 971–981 40 Amiry-Moghaddam, M. et al. (2005) Brain mitochondria contain aquaporin water channels: evidence for the expression of a short AQP9 isoform in the inner mitochondrial membrane. FASEB J. 19, 1459–1467 41 Rojek, A.M. et al. (2007) Defective glycerol metabolism in aquaporin 9 (AQP9) knockout mice. Proc. Natl. Acad. Sci. U. S. A. 104, 3609–3614 42 Manley, G.T. et al. (2000) Aquaporin-4 deletion in mice reduces brain edema after acute water intoxication and ischemic stroke. Nat. Med. 6, 159–163 43 Papadopoulos, M.C. and Verkman, A.S. (2005) Aquaporin-4 gene disruption in mice reduces brain swelling and mortality in pneumococcal meningitis. J. Biol. Chem. 280, 13906–13912 44 Amiry-Moghaddam, M. et al. (2004) a-syntrophin deletion removes the perivascular but not endothelial pool of aquaporin-4 at the blood-brain barrier and delays the development of brain edema in an experimental model of acute hyponatremia. FASEB J. 18, 542–544 45 Amiry-Moghaddam, M. et al. (2003) An a-syntrophin-dependent pool of AQP4 in astroglial end-feet confers bidirectional water flow between blood and brain. Proc. Natl. Acad. Sci. U. S. A. 100, 2106–2111 46 Papadopoulos, M.C. et al. (2004) Aquaporin-4 facilitates reabsorption of excess fluid in vasogenic brain edema. FASEB J. 18, 1291–1293 47 Tsuyumu, M. et al. (1981) Dynamics of formation and resolution of vasogenic brain oedema. I. Measurement of oedema clearance into ventricular CSF. Acta Neurochir. (Wien) 57, 1–13

48 Reulen, H.J. et al. (1978) Clearance of edema fluid into cerebrospinal fluid. A mechanism for resolution of vasogenic brain edema. J. Neurosurg. 48, 754–764 49 Bloch, O. et al. (2006) Accelerated progression of kaolin-induced hydrocephalus in aquaporin-4 deficient mice. J. Cereb. Blood Flow Metab. 26, 1527–1537 50 Greitz, D. et al. (1997) A new view on the CSF-circulation with the potential for pharmacological treatment of childhood hydrocephalus. Acta Paediatr. 86, 125–132 51 Saadoun, S. et al. (2005) Impairment of angiogenesis and cell migration by targeted aquaporin-1 gene disruption. Nature 434, 786–792 52 Papadopoulos, M.C. et al. Aquaporins and cell migration. Pflugers Arch. (in press) 53 Fawcett, J.W. and Asher, R.A. (1999) The glial scar and central nervous system repair. Brain Res. Bull. 49, 377–391 54 Yang, B. et al. (1996) The mercurial insensitive water channel (AQP-4) forms orthogonal arrays in stably transfected Chinese hamster ovary cells. J. Biol. Chem. 271, 4577–4580 55 Moon, C. et al. (2003) Involvement of aquaporins in colorectal carcinogenesis. Oncogene 22, 6699–6703 56 Hoque, M.O. et al. (2006) Aquaporin 1 is overexpressed in lung cancer and stimulates NIH-3T3 cell proliferation and anchorage-independent growth. Am. J. Pathol. 168, 1345–1353 57 Hu, J. and Verkman, A.S. (2006) Increased migration and metastatic potential of tumor cells expressing aquaporin water channels. FASEB J. 20, 1892–1894 58 Nagelhus, E.A. et al. (1999) Immunogold evidence suggests that coupling of K+ siphoning and water transport in rat retinal Muller ¨ cells is mediated by a coenrichment of Kir4.1 and AQP4 in specific membrane domains. Glia 26, 47–54 59 Binder, D.K. et al. (2004) Increased seizure threshold in mice lacking aquaporin-4 water channels. Neuroreport 15, 259–262 60 Binder, D.K. et al. (2006) Increased seizure duration and slowed potassium kinetics in mice lacking aquaporin-4 water channels. Glia 53, 631–636 61 Ding, J.H. et al. (2007) Alterations of striatal neurotransmitter release in aquaporin-4 deficient mice: An in vivo microdialysis study. Neurosci. Lett. 422, 175–180 62 Padmawar, P. et al. (2005) K+ waves in brain cortex visualized using a longwavelength K+-sensing fluorescent indicator. Nat. Methods 2, 825–827 63 Zhang, H. and Verkman, A.S. Aquaporin-4 independent Kir4.1 K+ channel function in brain glial cells. Mol. Cell. Neurosci. (in press) 64 Ruiz-Ederra, J. et al. (2007) Evidence against functional interaction between aquaporin-4 water channels and Kir4.1 potassium channels in retinal Muller cells. J. Biol. Chem. 282, 21866–21872 ¨ 65 Binder, D.K. et al. (2004) In vivo measurement of brain extracellular space diffusion by cortical surface photobleaching. J. Neurosci. 24, 8049–8056 66 Zador, Z. et al. Microfiberoptic fluorescence photobleaching reveals size-dependent macromolecule diffusion in extracellular space deep in brain. FASEB J. (in press) 67 Pellerin, L. et al. (2007) Activity-dependent regulation of energy metabolism by astrocytes: an update. Glia 55, 1251–1262 68 Yang, B. et al. (2006) Evidence against functionally significant aquaporin expression in mitochondria. J. Biol. Chem. 281, 16202–16206 69 Klatzo, I. (1994) Evolution of brain edema concepts. Acta Neurochir. Suppl. (Wien) 60, 3–6

Free journals for developing countries
In 2002, the WHO and six medical journal publishers launched the Health InterNetwork Access to Research Initiative, which enabled nearly 70 of the world’s poorest countries to gain free or reduced-cost access to biomedical literature through the internet. Currently more than 70 publishers are participating in the program, providing access to over 2000 journals. Gro Harlem Brundtland, former director-general for the WHO, said that this initiative was ‘‘perhaps the biggest step ever taken towards reducing the health information gap between rich and poor countries’’.

For more information, visit www.who.int/hinari
www.sciencedirect.com

Sign up to vote on this title
UsefulNot useful