Abstract

Introduction

Membrane transporters of the slc4 and slc9 family are expressed in many tissues mediating numerous functions including acid-base regulation and movement of large amounts of fluid and solutes.

The Na⁺ dependent acid-base transporters, slc4a5, −7, −10, and −11, and slc9a1 are all expressed in the choroid plexus epithelium (CPE) (Praetorius et al., 2004a; Bouzinova et al., 2005; Damkier et al., 2007; Damkier and Praetorius, 2012) where they, in addition to intra- and extracellular acid-base regulation, are involved in cerebrospinal fluid (CSF) secretion.

Over the last decades many genetically modified mouse models have been developed targeting these transporters. The phenotypes reported in some of these mice highlight the significance of the transporters in the CPE. The objective of the current review is to extract the insights gathered from these studies to give an overview of the important role of Na⁺ dependent pH regulators in the CPE. Removal of a protein by genetic modification in animal models gives an indication of the consequences of genetic mutations in human. It does, however, not necessarily provide insight into the physiological role of the transporter in the tissue. An important lesson from the studies of knockout mice is the compensatory mechanisms that take place when removing a protein by genetic modification. This makes it difficult to extrapolate the role of the isolated protein in the tissue. This role is potentially not the same as that observed when inhibiting the protein pharmacologically. In this review we will attempt to extract the data from the knockout mice and compare these to the human mutations and pharmacological studies of these transporters in the CPE and finally relate to studies of the same transporters in other epithelia.

Choroid plexus structure and morphology

The CPE constitutes a specialized and important part of the blood-CSF barrier (BCSFB). This barrier consists of endothelium, endothelial basement membrane, interstitial space with sparse connective tissue, subepithelial basement membrane and a single layer of cuboidal epithelial cells (Figure ​(Figure11).

During fetal development, certain areas of the brain ependyma become highly vascularized (Zagorska-Swiezy et al., 2008) and form the specialized secretory epithelium, protruding into the ventricles. The CPE is found in all four ventricles of the brain, forming branched, villous structures.

In contrast to the vasculature that constitutes the tight blood-brain barrier, the capillaries of the CPE are fenestrated and thus the fluid in the interstitial space resembles plasma ultra-filtrate. In fact, molecules as large as ferritin (~450kDa) are able to pass across the endothelium (Hurley et al., 1981). The permeability of the CP vasculature may be subject to regulation, as vascular endothelial growth factor secreted by CP epithelial cells both maintain (Kamba et al., 2006; Maharaj et al., 2008) and induce (Esser et al., 1998) the formation of fenestrations in endothelium.

The barrier function of the BCSFB is therefore mainly generated through the tightness of the CPE, as the single layer of cells is sealed by tight junctions (Wolburg et al., 2001). This is in contrast to the blood-brain barrier where the barrier function is made up of the tight junctions in the endothelial cells of the vascular wall (Figure ​(Figure3).3). The basolateral membrane of the CPE is highly convoluted and intertwining, allowing for rapid transport of solutes to and from the cytoplasm. The apical membrane of the epithelium is covered with microvilli, enlarging the cell surface area and facilitating the role of the epithelium as a highly secretory organ and regulator of CSF composition. The CPE cells also contain clusters of cilia (Mestres et al., 2011), critical for maintaining the flow and possibly the secretion of CSF (Narita et al., 2010).

Cerebrospinal fluid secretion

The main role of the CPE is to produce CSF. The majority (approximately 80%) of the CSF is secreted by the CPE while the remaining 20% is derived from the brain interstitial fluid (Redzic et al., 2005). The adult human CPE secretes approximately 0.5 L CSF/day at a rate of approximately 0.2–0.4 ml/min/g tissue. The total CSF volume surrounding the brain and spinal cord is estimated to be about 150 ml in adults and is distributed in the cranial and spinal subarachnoid spaces (125 ml) and in the brain ventricles (25 ml) (Sakka et al., 2011). The [Na⁺]_CSF in rabbit was measured to be 149 mM compared to 148 mM in plasma. The [HCO₃⁻]_CSF in rabbit was estimated to 22 mM compared to 25 mM in plasma (Davson and Segal, 1996). The composition of the ions in the CSF differs from that predicted for a plasma ultrafiltrate as Na⁺, Mg⁺⁺, and HCO₃⁻ concentrations are higher and K⁺ and Ca⁺⁺ are lower in CSF (Ames et al., 1964). Thus, the CSF is actively secreted by the CPE and the concentrations of these ions stay relatively constant despite fluctuations in plasma (Husted and Reed, 1977; Murphy et al., 1986). Finally, CSF is slightly hypertonic compared to plasma (approximately 5 mM CSF positive, Davson and Purvis, 1954) which would not be expected from an ultra-filtrate.

CSF secretion by the CPE is unlike in other secretory epithelia driven by an osmotic gradient generated by the transcellular movement of Na⁺ from blood through CPE to the CSF (Wright, 1972). In other secretory epithelia, Cl⁻ drives secretion. The osmotic gradient drives water to be moved from plasma to the CSF. Unlike most other secretory epithelia the Na, K ATPase is located in the luminal membrane of the CPE (Figure ​(Figure2).2). The Na, K ATPase is pivotal for creating the Na⁺ gradient that drives Na⁺ import across the basolateral membrane (Ames et al., 1965). Na⁺ enters the cell presumably via the Na⁺ dependent HCO₃⁻ co-transporter, NCBE. The movement of water through the CPE is mainly believed to be mediated by AQP1 located in the luminal membrane and to a smaller degree in the basolateral membrane (Figure ​(Figure1)1) (Nielsen et al., 1993; Praetorius, 2007). Some water-movement could also occur through a paracellular route, e.g., via claudin-2 (Rosenthal et al., 2010).

Regulation of pH in brain and CSF

Acid-base balance in the extracellular fluid surrounding the brain is a prerequisite for optimal brain function, as the neurons are very sensitive to changes in pH (Balestrino and Somjen, 1988). The blood brain barrier (BBB) is virtually impermeable to HCO₃⁻ and H⁺ (De Bersaques and Leusen, 1954; Siesjo and Ponten, 1966) while CO₂ readily diffuses across the BBB into the brain parenchyma (Ponten, 1964; Geiser et al., 1996). CO₂ from the blood is in equilibrium with the CO₂ in the CSF, and numerous studies have shown that changes in arterial pCO₂ affect CSF pCO₂, pH, and ventilation (Pappenheimer et al., 1965; Fencl et al., 1966; Loeschcke and Sugioka, 1969; Andrews et al., 1994). Furthermore, it is known that neuronal excitability is enhanced by increases in brain pH, and decreased when brain pH is lowered (Balestrino and Somjen, 1988). A pioneering study showed that rat pups exposed to hyperthermia developed respiratory alkalosis and febrile seizures (Schuchmann et al., 2006). Seizures have long been known to be halted by inhalation of CO₂ (Pollock et al., 1949). CSF is practically devoid of proteins, meaning that the non-HCO₃⁻ buffering power is almost non-existing, and, as the BBB is impermeable to HCO₃⁻ and H⁺, the brain must rely on other mechanisms for regulating its acid-base balance in the face of changes in plasma pCO₂. Apart from the ventilatory response upon changes in CSF/extracellular acid-base status, it is believed that the CPE plays a role in the regulation of CSF pH by secretion and absorption of acid/base equivalents. In 1967, Kazemi showed an almost identical increase in arterial blood pCO₂ and CSF pCO₂ in dogs after 1 h of 10% CO₂ inhalation. In the dogs that hyperventilated, similar drops in pCO₂ were observed in blood and CSF. The ensuing changes in CSF pH were of the same magnitude as the changes in blood pH in both experiments after 1 h. However, 6 h into the experiment, CSF pH had returned toward its resting value, whereas no similar compensation was observed in the blood (Kazemi et al., 1967). This indicates that CSF pH is somehow protected against long-term acid-base disturbances. Furthermore, CO₂-stimulated increases in CSF [HCO₃⁻] are accompanied by an increase in CSF [Na⁺] (Nattie, 1980), indicating that buffering of CSF pH is Na⁺-dependent. As the CPE expresses a range of Na⁺ dependent acid-base transporters (Figure ​(Figure2)2) it is therefore well equipped for protecting CSF pH against the effects of changes in blood pCO₂.

Basolateral Na⁺ dependent acid-base transporters

The Na⁺ dependent Cl⁻/HCO₃⁻ exchanger, NCBE (slc4a10), is expressed in the basolateral membrane of the CPE (Figures ​(Figures1,1, ​,2)2) (Praetorius et al., 2004a), where it performs Na⁺-dependent Cl⁻/HCO₃⁻ exchange (Wang et al., 2000; Damkier et al., 2010a), although the precise mode of action differs between expression systems (Parker et al., 2008). In all cases NCBE mediates the transport of Na⁺ and HCO₃⁻ into the cell. In rat and human the electroneutral Na⁺:HCO₃⁻ co-transporter, NBCn1 is also found in the basolateral membrane of the CPE (Praetorius and Nielsen, 2006). The Na⁺ dependent Cl⁻/HCO₃⁻ exchanger (slc4a8) is expressed in the basolateral membrane of the rat fetus (Chen et al., 2008a), however, in adult rats the expression is absent suggesting a role of NDCBE in choroid plexus development.

Luminal Na⁺ dependent acid-base transporters

The slc4a5 gene product, NBCe2, is expressed in the luminal membrane of the CPE (Figure ​(Figure2;2; Bouzinova et al., 2005). NBCe2 mediates the transport of three HCO₃⁻ together with one Na⁺ from the cell to the CSF. This stoichiometry suggests a role of NBCe2 in CSF pH regulation (Millar and Brown, 2008). In mouse, and in some cases in humans, NBCn1 (slc4a7) is expressed in the luminal membrane (Praetorius and Nielsen, 2006). The apparent discrepancy in localization of this protein between species could indicate that the protein in CPE is most likely involved in protecting the cell from intracellular acidification rather than in secretion of CSF as the direction of transport regardless of the membrane expression is most likely inwards. The Na⁺ Borate transporter (slc4a11) is similarly expressed in the luminal membrane of CPE (Damkier et al., 2007). This transporter was recently shown to transport Na⁺ in exchange for H⁺ suggesting a similar role as the Na⁺/H⁺ exchanger, NHE1, in the luminal membrane. Finally, NHE1 (slc9a1) is present in the luminal membrane of the CPE (Damkier et al., 2009). The localization of NHE1 indicates that this protein, similar to NBCe2, could be an important mechanism for regulating CSF pH.

In addition to these transporters the CPE expresses numerous other transporters involved in the secretion of solutes and nutrients. These are outside the scope of this review but are reviewed elsewhere (Damkier et al., 2010b; Ho et al., 2012).

Choroid plexus as target for pharmacological treatment

Delivery of drugs from blood to the brain is greatly limited by the tightness of the barriers in the brain, the BBB and the BCSFB. The CPE is the key component in the BCSFB. The physical tightness of the BCSFB is accomplished by the tight junctions connecting the epithelial cells, whereas the tightness of the larger BBB is made by the tight junctions between the endothelial cells in the brain capillaries (Figure ​(Figure3;3; Reese and Karnovsky, 1967). In addition to the physical tightness, both the BCSFB and especially the BBB contain ATP binding cassette (ABC) transporters that actively rid the brain of xenobiotics thereby further preventing drug-delivery to the brain. Therapeutic strategies have targeted these transporters to increase delivery to the brain through the BBB (Hartz and Bauer, 2010). In some cases, injections of therapeutics directly into the subarachnoid space (intrathecal administration) are used as targeted treatments of diseases in the brain such as tumors and infections (Varelas et al., 2008; Serwer and James, 2012).

As the physical barrier function of the BCSFB is not in the vasculature but rather in the CPE cells the proteins in the basolateral membrane of CPE could potentially be targeted in therapy to inhibit CPE functions, e.g., secretion and pH regulation of CSF. Given the dramatic effects on CSF secretion of knocking out NCBE, this protein is an obvious drug target for reducing CSF secretion. Direct neuronal side effects of inhibition of NCBE may not be relevant since the protein within the CNS would be protected by the BBB and thereby only the CPE would be targeted. However, brain pH could be affected through equilibration with a possibly altered CSF pH. The brain is covered by the rigid skull which necessitates a tight regulation of volume and thereby intracranial pressure (ICP) in order to avoid damage to the brain tissue. Sudden increases in ICP can in the worst case lead to fatal incarceration. Overall, increases in volume inside the skull can be placed within the three “compartments” of the brain; the brain tissue (tumors), the blood (hemorrhage, swelling following ischemia or trauma) or the CSF (hydrocephalus). According to the Monro-Kelli doctrine, any increase in one of these compartments needs to be balanced by a decrease in another (Mokri, 2001). Decreasing CSF secretion during the acute stages of increases in ICP could be extremely beneficial for the patient outcome.

Regulation of CSF pH and thereby brain pH is essential for normal brain function and is similarly important for inhibition of seizures. Both the NCBE and NBCe2 knockout mice seem to have a higher seizure threshold than wild type. It is known that lowering of pH during a seizure causes an activation of the acid sensing Na⁺ channel, ASIC1a, leading to an inhibition of neuronal firing and thereby cessation of the seizure (Ziemann et al., 2008). The loss of a bicarbonate transporter in the NBCe2 and NCBE knockout mice most likely causes a decrease in CSF-pH which could potentially be protective for the brain. In addition to inhalation of CO₂ in the acute stages of a seizure these proteins could be potential targets in the long-term treatment of seizure disorders by acutely lowering CSF pH and thereby brain pH.

Conclusions

The study of choroid plexus physiology by use of knockout mice targeting the Na⁺ dependent acid-base transporters of the slc4 and slc9 families have so far highlighted many aspects of the significance of these transporters in CSF secretion and pH regulation. Further studies are warranted for highlighting the role of slc4a11 and slc9a1 in choroid plexus biology.

The studies highlight the importance of epithelial transporters in vivo, however the findings in genetically modified mice include the compensatory mechanisms that take place during fetal development and early life, and call for a general need for caution when interpreting these data. The genetic removal of some transporters, such as NCBE and NBCe2 leads to restructuring of the entire epithelium. This is interesting from a cell structure point-of-view but when investigating the particular protein of interest these compensatory changes could also completely change cell function and are thus not specific to removal of one protein. The compensatory changes could potentially be minimized by either creating an inducible knockout mouse allowing the mouse to fully develop before inducing the deletion, by generating mice that carry a specific mutation that blocks the NBC action, or by intraventricular installation of siRNA targeting the gene product. These kinds of studies introduce other complicating factors including the success rate of knockdown. This emphasizes the need for selective inhibitors both for studying the transporters of the slc4 family but also for use in the pharmacological targeting of the transporters in disease.

Author contributions

The manuscript was written with equal contributions from all authors. Final editing was mainly carried out by Henriette L. Christensen and Helle H. Damkier. Figures were created by Helle H. Damkier.

Acknowledgments

References