Increased Aquaporin-1 and Na1-K1-2Cl–

Cotransporter 1 Expression in ChoroidPlexus Leads to Blood–Cerebrospinal FluidBarrier Disruption and Necrosis ofHippocampal CA1 Cells in Acute RatModels of Hyponatremia

Jaehyun Kim1 and Yongwook Jung1,2*1Department of Anatomy, College of Medicine, Dongguk University, Gyeongju, Republic of Korea2Section of Neuroscience Research, Medical Institute of Dongguk University, Gyeongju,Republic of Korea

Hyponatremia is a metabolic disorder characterized byincreased cerebrospinal fluid (CSF) volume and pressure,although the site of brain insult is unclear. Specifically,the hippocampus, which is in direct contact with expand-ing CSF ventricles, may be involved. The present studywas undertaken to investigate the possible roles of cho-roid plexus aquaporin-1 (AQP1) and of cation chloridetransporters (Na1-K1-2Cl– cotransporter 1 [NKCC1] andK1-Cl– cotransporter 4 [KCC4]) in the underlying hippo-campal pathophysiology of hyponatremia in acute (6 and12 hr duration) experimental models. It was found thatthe expressions of AQP1 and NKCC1 proteins in choroidplexus were significantly increased, whereas the expres-sion of KCC4 protein was unchanged vs. control valuesafter 6 and 12 hr of hyponatremia. Choroid plexuses withincreased AQP1 and NKCC1 after 6 hr of hyponatremiashowed caspase 3-dependent apoptosis and disruptionof the blood–CSF barrier. Furthermore, necrotic changesin CA1 neuronal cells were observed after 6 and 12 hr ofhyponatremia. Overall, these data suggest that increasesin AQP1 and NKCC1 expression under hyposmoticstress may be one of the molecular mechanisms underly-ing the pathophysiology of acute hyponatremia, such asthe necrotic cell death of hippocampal CA1 region byincreasing water transport across the blood–CSF barrier.Furthermore, we suggest that opening of the blood–CSFbarrier after acute hyponatremia may be triggered thesecondary adverse conditions that are capable ofenhancing selective necrosis in hippocampal CA1cells. VVC 2012 Wiley Periodicals, Inc.

Key words: acute hyponatremia; AQP1; NKCC1;hippocampus

Patients with acute hyponatremia (defined as hypo-natremia of duration <12 hr) have a 29% seizure inci-dence and a mortality rate of 50%, whereas patients with

chronic hyponatremia (� 2–3 days duration) have 4%seizure incidence and a mortality rate of 6% (Sterns andSilver, 2006). However, in vivo experiments in rats haveproduced contrasting results and have led to some doubtabout the molecular mechanism underlying the neuronaldeath responsible for acute hyponatremic stress(Pasantes-Morales and Tuz, 2006; Ayus et al., 2008;Hyzinski-Garcia et al., 2011).

The choroid plexus is a branched structure madeup of a single layer of epithelial cells and fenestratedblood capillaries, which are unlike those in the blood–brain barrier (BBB). Furthermore, choroid plexus epi-thelial cells form a blood–CSF barrier because of junc-tional complexes between the cells, which restrict thepassage of molecules and ions into the cerebrospinal fluid(CSF; Spector and Johanson, 1989; Johanson, 1993). Itis widely believed that alterations in CSF amounts areplausibly involved in perturbed neural function via thedysregulation of intracranial pressure (ICP; Melton andNattie, 1984).

Aquaporin-1 (AQP1) functions primarily as a waterpore and facilitates the transmembrane transport of waterdriven by osmotic gradients. Furthermore, the apicallocalization of AQP1 in the choroid plexus suggests thatit plays an important role in facilitating water transportduring CSF secretion (Nielsen et al., 1993; Hasegawa

Contract grant sponsor: Korean Research Foundation; Contract grant

number: KRF-2009-013-E00032.

*Correspondence to: Yongwook Jung, MD, PhD, Department of Anat-

omy, College of Medicine, Dongguk University, Sukjang-dong 707,

Gyeongju 780-714, Republic of Korea. E-mail: jungyw@dongguk.ac.kr

Received 11 August 2011; Revised 21 November 2011; Accepted 2

December 2011

Published online in Wiley Online Library (wileyonlinelibrary.com).

DOI: 10.1002/jnr.23017

Journal of Neuroscience Research 00:000–000 (2012)

' 2012 Wiley Periodicals, Inc.

et al., 1994). Therefore, the concentration of AQP1 inthe epithelial cells of the apical membrane determinesCSF secretion. Recently, Liu et al. (2010) found thatAQP1 siRNA efficiently and specifically inhibited theproliferation and induced apoptosis of K562 cells. Otherstudies have proposed that AQP1 is involved in themovement of water across the plasma membrane indying cells during apoptotic volume decrease (AVD;Jablonski and Hughes, 2006).

Cation-chloride cotransporters are a family of seventransporters that mediate electroneutral Cl– and cation(Na1 and/or K1) transport in the same direction acrossthe cell membrane (Hebert et al., 2004). Na1-K1-2Cl–

cotransporter 1 (NKCC1) is expressed in the apicalmembrane of choroid plexus epithelial cells and isbelieved to mediate ion efflux because choroid plexuscells contain higher levels of Na1 and Cl– than othercell types (Plotkin et al., 1997). Recently, Poulsen et al.(2010) reported that the protection of multidrug-resistantEhrlich ascites tumor cells (EATC) against apoptosisinvolves the attenuation of water loss during AVD likelyvia reduced Cl– loss. The importance of Cl– efflux wasalso highlighted by the observations that the NKCC1blocker bumetadine reduced the water loss, cell shrink-age, and caspase 3 activation in EATC. A recent immu-nohistochemical study showed that K1-Cl– cotransporter4 (KCC4) protein is expressed in the apical membraneof mouse choroid plexus. The role of KCC4 in the api-cal membrane is less obvious, but it might contribute toCl– transport into the CSF and to the recycling of K1,which occurs at this membrane (Karadshen et al., 2004).

The aims of the present study were to identifychanges in the expression and function of choroidalAQP1, NKCC1, and KCC4 in rodent models of acutehyponatremia. In addition, we evaluated the detrimentaleffects of altered expressions of these channels in thepathogenesis of the hippocampus. The hippocampus hasbeen chosen in investigations of hyponatremia-inducedbrain damage because of its close proximity to the ven-tricular CSF.

MATERIALS AND METHODS

Induction of Hyponatremia

The studies were carried out on 21 adult male Sprague-Dawley rats (250–280 g) with free access to drinking waterand standard rodent food pellets. Hyponatremia was inducedas described by Vajda et al. (2000), with some modification.Acute hyponatremia was induced by simultaneous water load-ing (140 mM/liter dextrose solution) and administering 8-dea-mino-arginin vasopressin (dDAVP) in saline and DMSO (30%v/v): 6 hr group 30 ml (� 12% body weight) dextrose solu-tion i.p. and 3 lg dDAVP s.c. followed by repeated doses of20 ml (� 8% bw) dextrose i.p. and 3 lg dDAVP s.c. at 4 hr;12 hr group 30 ml (� 12% bw) dextrose solution solution i.p.and 3 lg dDAVP s.c., followed by repeated doses of 20 ml(� 8% bw) dextrose i.p. and 3 lg dDAVP s.c. at 4 and 8 hr.Animals were sacrificed 6 and 12 hr after the first injection.After completing the hydration protocols, rats were anesthe-

tized with isoflurane; venous blood was taken to determineserum osmolarity and sodium concentration; and the brainwas removed for Western blotting, TUNEL staining, immu-nohistochemistry, and cresyl violet studies. The experimentalprocedures were reviewed and approved by the Animal Careand Use Committee of Dongguk University. Animal care anduse were in accordance with the guidelines issued by theNational Institutes of Health.

Immunohistochemistry

For immunohistochemistry, brains were fixed using atranscardiac infusion of 4% paraformaldehyde. Perfused brainswere then removed and paraffin embedded. Serial coronal 5-lm sections were obtained at the level of the dorsal third ven-tricle (bregma –4.30 mm). After deparaffinization, sectionswere stained using a Dako kit (Santa Cruz Biotechnology,Santa Cruz, CA). Endogenous peroxidase activity was blockedby incubating sections for 5 min in 3.0% H2O2. After washingin phosphate buffer solution (PBS), sections were incubatedwith rabbit anti-rat AQP1 (Chemicon, Temecula, CA; 1:200)and rabbit anti-IgG (Chemicon; 1:500) for 16–18 hr at 48Cand then with biotinylated universal anti-mouse, -goat, and -rabbit immunoglobulins in PBS for 30 min. After washing inPBS, sections were incubated with streptoavidin conjugated tohorseradish peroxidase (HRP) in PBS for 30 min. Finally,they were reacted with a solution containing diaminobenzi-dine (DAB) and hydrogen peroxide (0.001%), and then coun-terstained with Mayer’s hematoxylin to visualize cell nuclei.

TUNEL Staining

TUNEL staining (ApopTag kit; Intergen) was employedfor the detection of DNA fragmentation in the cells of cho-roid plexus and hippocampus. Briefly, after digesting proteinusing proteinase K and quenching endogenous peroxidase ac-tivity with 3.0% H2O2 in PBS, slides were placed in equilibra-tion buffer and then in working TdT enzyme, followed bystop/wash buffer. After applying two drops of antidigoxige-nin-peroxidase to slides, peroxidase was detected with DAB.Negative controls were prepared using distilled water or PBSinstead of TdT enzyme when preparing working TdT.

Cresyl Violet Staining

After being washed in distilled water, sections werestained with cresyl violet for 30 min at RT and then treatedsuccessively with ethanol (50%, 70%, 95%, 100%) and differ-entiator (glacial acetic acid and 95% ethanol).

SDS-PAGE and Immunoblotting

Choroid plexuses were removed from hyponatremic rats(n 5 8) and control rats (n 5 8). Tissue samples were ho-mogenized in homogenizing buffer (0.32 M sucrose, 25 mMimidazole, 1 mM EDTA, pH 7.2, containing 8.5 mM leupep-tin, 1 mM phenylmethylsulfonyl fluoride) for 10 sec using aPolytron. Aliquots were stored at –708C. Samples of homoge-nate were run on 9–15% polyacrylamide minigels (Bio-RadMini Protean II; Bio-Rad, Hercules, CA). For each gel, anidentical gel was run in parallel and Coomassie blue stained toensure identical loading, and the other gel was immuno-

2 Kim and Jung

Journal of Neuroscience Research

blotted. After electrophoresis, proteins were transferred tonitrocellulose membrane in a buffer solution containing 50mM Tris-base, 380 mM glycine, and 20% methanol. Mem-branes were blocked with 5% milk in PBS-T (80 mMNa2HPO4, 20 mM NaH2PO4, 100 mM NaCl, 0.1% Tween20, pH 7.5) for 1 hr and incubated overnight at 48C with rab-bit anti-AQP1 (Chemicon; 1:1,000), rabbit anti-NKCC1(Chemicon; 1:500), rabbit anti-KCC4 (Santa Cruz Biotech-nology, Santa Cruz, CA; 1:500), and mouse anticaspase 3 (L-18; Santa Cruz Biotechnology; 1:1,000). The sites of anti-body–antigen reaction were visualized with horseradish perox-idase (HRP)-conjugated secondary antibodies (P447 or P448,diluted 1:3,000; Dako, Glostrup, Denmark) and an enhancedchemiluminescence system (ECL; Amersham Pharmacia Bio-tech, Little Chalfont, United Kingdom) and exposed to pho-tographic film (Hyperfilm ECL; RPN3103K; AmershamPharmacia Biotech). The immunoblot signals developed bythe ECL system were quantified in Scion Image (version1.59).

Presentation of Data and Statistical Analyses

Quantitative data are presented as means 6 standarderrors of mean (SEMs). Comparisons between groups weremade using the unpaired Student’s t-test. P < 0.01 or P <0.05 was considered significant.

RESULTS

Establishment of Acute Hyponatremia

In agreement with a previous study (Vajda et al.,2000), the combined administration of hypotonic dex-trose solution and dDAVP induced a progressive fall inthe concentrations of serum Na1 and serum osmolarity:117 6 2 mM and 252 6 6 mOsm/liter (for a 6-hr du-ration) and 111 6 5 mM and 245 6 6 mOsm/liter (fora 12-hr duration) compared with 140 6 4 mM and 2966 5 mOsm/L, respectively, in controls. Thus, a notablereduction in intravascular osmolarity would promotewater transport from plasma into the brain.

Altered Expressions of AQP1, NKCC1, and KCC4in Choroid Plexus in Response to AcuteHyponatremia

AQP1 was localized in choroid plexus samples pre-pared from control and hyponatremic rats using immuno-histochemical methods based on anti-AQP1 antibody.Distinct immunoreactive AQP1 signals were observedmainly in the apical membrane of the third ventricle cho-roid plexus epithelium in controls (Fig. 1A, arrow). Im-munohistochemical examination of AQP1 in choroidplexus demonstrated that its immunoreactivity wasincreased by 6 and 12 hr of hyponatremia (Fig. 1B,C). Toevaluate quantitatively the effect of hyponatremiaon AQP1 protein expression, immunoblotting using anti-AQP1 antibody was performed on third-ventricle choroidplexus, and this revealed a major, strong band at 28 kDa.b-Tubulin was used as a loading control (Fig. 1D). Con-sistent with immunohistochemical findings, immunoblot-ting analysis showed that AQP1 expression in choroid

plexus after hyponatremia was three to five times higherthan in controls (6 hr: 318% 6 7%; 12 hr: 531% 6 4.0%of control, n 5 4, P < 0.01; Fig. 1E). Immunoblottingwith NKCC1 and KCC4 antibodies revealed products of110 and 135 kDa, respectively. a-Actin was used as inter-nal control for confirming equal loading of proteins (Fig.2A). Consistent with AQP1 expression, NKCC1 expres-sion was significantly increased after 6 and 12 hr of hypo-natremia (6 hr: 133% 6 4%; 12 hr: 131% 6 12% of con-trol, n 5 3, P < 0.05; Fig. 2B). However, the expressionof KCC4 in choroid plexus was unchanged in response tohyponatremia.

Caspase 3-Dependent Apoptosis in Choroid Plexusand Disruption of the Blood–CSF Barrier inResponse to Acute Hyponatremia

No TUNEL-positive epithelial cells were observedin control animals (data not shown). In contrast to thecase in control animals, TUNEL-positive cells (blackarrows) with small and condensed nuclei and intactplasma membrane were dramatically higher after 6 hr of

Fig. 1. Expression of aquaporin1 (AQP1) in rat choroid plexus increased af-ter 6 and 12 hr of acute hyponatremia. A: Representative immunohisto-chemical image of a choroid plexus from a control rat demonstrating thatAQP1 selectively localizes on the apical surface of the choroid plexus. Thearea indicated by the arrow is magnified in the inset. AQP1 immunoreactiv-ity after 6 (B) and 12 (C) hr of hyponatremia was stronger than in controls,indicating an increase in AQP1 expression. D: Immunoblot bands reactedwith anti-AQP1 antibody after 6 and 12 hr of hyponatremia. b-Tubulinwas used as a loading control. E:Densitometric analysis of AQP1 expression.Normalized expression levels of AQP1 after 6 and 12 hr of hyponatremiawere significantly increased. Data are means6 SEMs. Control n5 4, hypo-natremia n 5 8. *P < 0.01. Scale bar 5 50 lm. [Color figure can beviewed in the online issue, which is available at wileyonlinelibrary.com.]

AQP1 and NKCC1 in Choroid Plexus 3

Journal of Neuroscience Research

hyponatremia and intermingled with TUNNEL-negativecells, which are light and consequently not DAB stained(Fig. 3A). Thereafter, TUNEL-positive cells almost disap-peared after 12 hr of hyponatremia (data not shown).

Because epithelial cells in the choroid plexus constitutethe blood–CSF barrier, we tested whether the blood–CSF barrier was disrupted. We checked the localizationof plasma immunoglobin G (IgG) in ventricular space todetermine whether blood–CSF barrier disruption wasinduced by the apoptosis of choroid epithelial cells.Extensive positive staining of extravasated plasma IgG wasfound within the third ventricle after 6 hr of hyponatre-mia (Fig. 3B), indicating blood–CSF barrier impairment.However, extravasation of plasma IgG into the ventriclewas not observed after 12 hr of hyponatremia (data notshown). Caspase 3 is a key executor of apoptosis (Poulsenet al., 2010). To determine whether TUNEL-positivecells underwent apoptosis in choroid plexus, we investi-gated the presence of cleaved procaspase 3 by immuno-blotting. Immunoblotting for procaspase 3 (CPP32)revealed a major, strong band at 32 kDa in controls. a-Actin was used as internal control for confirming equalloading of proteins (Fig. 3C). A significant reduction inCPP32 after 6 hr of hyponatremia (86% 6 2% of thecontrol level, n 5 3, P < 0.05) indicated the activationof caspase 3 and its active involvement in apoptotic proc-esses occurring in choroid plexus epithelial cells (Fig. 3D).

Necrotic Changes in Hippocampal CA1 Cells AfterAcute Hyponatremia

We used cresyl violet staining as a means of differ-entiating apoptotic and necrotic changes, because theTUNEL assay is incapable of doing so. CA1 and CA3subpopulations of hippocampal cells were chosen tocharacterize cellular changes under hyponatremic condi-tions. No apoptotic or necrotic cells were observed incontrol animals (Fig. 4A,D). However, loss of plasmamembrane integrity was observed in most CA1 cellswith cresyl violet staining after 6 hr of hyponatremia(Fig. 4B, black arrows). Thereafter, karyorrhexis wasprominent in the CA1 cells after 12 hr of hyponatremia

Fig. 3. Apoptosis of choroid plexus epithelial cells after 6 hr of acute hy-ponatremia as indicated by TUNEL staining. A: TUNEL-positive cells(arrows) with small and condensed nuclei and intact epithelial appearancewere increased after 6 hr of hyponatremia and intermingled with TUN-NEL-negative cells, which are light and consequently not DAB stained.B: Extravasated plasma IgG was seen in the third ventricle after 6 hr ofhyponatremia. C: Cleavage of procaspase 3 (CPP32) by immunoblotting.

a-Actin was used as internal control for confirming equal loading of pro-teins. D: A significant decrease in the normalized expression of CPP32was observed after 6 hr of hyponatremia, indicating caspase 3-dependentapoptosis in choroid plexus epithelial cells. Data are means 6 SEMs.Control n 5 3, hyponatremia n 5 6. *P < 0.05. Scale bar 5 50 lm.[Color figure can be viewed in the online issue, which is available atwileyonlinelibrary.com.]

Fig. 2. Expressions of Na1-K1-2Cl– cotransporter 1 (NKCC1) andK1-Cl– cotransporter 4 (KCC4) in rats with acute hyponatremia and incontrol rats. A: Immunoblot bands reacted with anti-NKCC1 and anti-KCC4 antibodies after 6 and 12 hr of hyponatremia. a-Actin was used asinternal control for confirming equal loading of proteins. B: Densitomet-ric analysis of NKCC1 and KCC4 expressions. Normalized expressionlevels of NKCC1 after 6 and 12 hr of hyponatremia were significantlyelevated whereas the expression of KCC4 was unchanged. Data aremeans 6 SEMs. Control n 5 4, hyponatremia n 5 6. *P < 0.05.

4 Kim and Jung

Journal of Neuroscience Research

(Fig. 4C, white arrows). Karyorrhexis occurs relativelylate during necrosis and involves the destructive frag-mentation of the nucleus of a dying cell, whereby itschromatin is distributed irregularly throughout the cyto-plasm (Majno and Joris, 1995). Therefore, mostTUNEL-positive cells in the CA1 region 6 hr after hy-ponatremia were the result of nucleotide binding to theDNA of necrotic cells rather than to oligonucleosomesof apoptotic cells (Fig. 4H, white arrow), although asmall number of TUNEL-positive cells had small andcondensed nuclei (Fig. 4H, black arrow). If CA1 deathhad occurred by apoptosis, most of the TUNEL-positivecells should have had small and condensed nuclei. Incontrast to the CA1 region, the labeling intensities ofTUNEL in CA3 cells decreased after 6 hr of hyponatre-mia (Fig. 4K, black arrow). Weak DAB-stained cells inthe TUNEL assay are generally considered false positive.This is caused by the release of endogenous endonucle-ases, and this process is highly dependent on the lengthof time for which sections are incubated with proteinaseK (Stahelin et al., 1998).

DISCUSSION

Role of Increased AQP1 and NKCC1 Eexpressionsin CSF Secretion Following Acute Hyponatremia

Expression of choroidal AQP1 was increased three-fold after 6 hr of hypoatremia, which indicates that thereis a massive increase in the osmotic water permeabilityof the blood–CSF barrier during hyponatremia. Previousstudy has demonstrated that a 1% decrease in plasma

osmolarity resulted in a 6.7% increase in CSF production(Di Mattio et al., 1975). It was suggested that the cho-roid plexus may account for only 25–50% of total CSFproduction (Milhorat et al., 1971). Additionally, osmoti-cally driven water permeability across the choroidal epi-thelium in AQP1 null mice was found to be reducedfivefold (Oshio et al., 2005). Thus, it is expected thatthe increase in AQP1 expression after 6 hr of hypoatre-mia could contribute approximately 5–10% increase tothe total CSF production. In theory, an increase in CSFvolume resulting from hyponatremia will increase ICPbecause the limits of expansion imposed by the skullprovide narrow margins for buffering intracranial volumechanges. Accordingly, as expansion occurs, compressingsmall vessels in the hippocampus generates episodes ofanoxic ischemia and subsequent neuronal death. In thepresent study, the expression of AQP1 was increasedfivefold after 12 hr of hyponatremia. However, casapase3-dependent apoptosis of choroid plexus developed after6 hr. One alternative, though more controversial, possi-bility was suggested by Jablonski et al. (2004), namely,that the function of AQP1 in thymocytes is inactivatedafter apoptosis. If this is correct, it is possible that, invivo, increased AQP1 in the choroid plexus after 12 hrof hyponatremia is inactivated by a posttranslationalmechanism. Therefore, we suggest that apoptosis of cho-roid plexus and subsequent inactivation of AQP1 pro-vide another defense mechanism for protecting the brainfrom extreme volume changes and that this provides areasonable explanation for tolerance to 12 hr of hypona-tremia, which is observed clinically.

Fig. 4. Necrosis in the hippocampal CA1 region was observed after6 and 12 hr of acute hyponatremia. A–C: Loss of plasma membraneintegrity was observed in most of CA1 cells after 6 hr (B, blackarrows). Thereafter, karyorrhexis was prominent in CA1 cells after12 hr by cresyl violet staining (C, white arrows). D–F: No changein CA3 cells was observed after 6 and 12 hr of hyponatremia bycresyl violet staining. G–I: Necrotic cells with TUNEL positivity in

CA1 after 6 hr showed large and noncondensed nuclei (H, whitearrow), even though a small number of TUNEL-positive cells hadsmall and condensed nuclei (H, black arrow). J–L: Most of the cellsin CA3 after 6 hr showed false TUNEL positivity compared withCA1. Control n 5 3, hyponatremia n 5 6. Scale bar 5 50 lm.[Color figure can be viewed in the online issue, which is availableat wileyonlinelibrary.com.]

AQP1 and NKCC1 in Choroid Plexus 5

Journal of Neuroscience Research

Consistent with the high expression of AQP1, ourresults demonstrate that NKCC1 was significantlyincreased after 6 hr of hyponatremia. NKCC1 is knownto promote water transport across cell membranes and toaid in the transcellular movement of Cl– across both se-cretory and absorptive epithelia (Payne and Forbush,1995). Studies performed on Xenopus oocytes transfectedwith NKCC1 have shown that NKCC1 works both as awater channel allowing passive water flux and as a waterpump that transports water across apical membranesregardless of osmotic gradients (Hamann et al., 2010).Recent studies have demonstrated that NKCC1 trans-ports 500 water molecules for each cycle of cation-chlo-ride transport (MacAulay and Zeuthen, 2010), which iscomparable to that of aquaporins. Moreover, the choroidplexus intracellular concentrations of Na1 and Cl– arevery high compared with most cell types, so the drivingforce for inward compared with outward transport ismuch reduced. Therefore, in the present study, the rela-tively high concentration of NKCC1 in the apical mem-brane of choroid plexus epithelium after 6 hr of hypona-tremia suggests that NKCC1 increases the rate of Na1,Cl–, and water efflux and that it may have an additionalrole in CSF oversecretion.

Our results also demonstrate that the expression ofNKCC1 was maintained after 12 hr of hyponatremia.However, as with AQP1, enhanced NKCC1 expressioncan trigger apoptosis (Iwamoto et al., 2004), because ap-optosis is associated with a significant reduction in cellvolume and significant losses of K1 and Na1 transportpathways. Recent studies have demonstrated that apo-ptosis-associated tyrosine kinase is increased during apo-ptosis (Gagnon et al., 2007) and that it indirectly inhibitsthe NKCC1 via protein phosphatase 1, which inactivatesthe NKCC1 activators, such as Ste-20-related proline-al-anine-rich kinase and with no lysine (K) kinase. There-fore, our data suggest that the inactivation of NKCC1activity via a dephosphorylation mechanism involves adefense mechanism for protecting the brain fromextreme volume and ICP changes.

Roles of Increased AQP1 and NKCC1 Expressionsin the Apoptosis of Choroid Plexus FollowingAcute Hyponatremia

Previous studies have demonstrated that the over-expression of AQP1 in Chinese hamster ovary cellsincreases the amount of apoptosis (Jablonski et al., 2004)and that hepatic tumor cells with inherent resistance toapoptotic stimuli show decreased expression of AQP8and -9 (Jablonski et al., 2007). Another study demon-strated that AQP1 knockdown may significantly inhibitthe activity of caspase 3 in human proximal tubular epi-thelial cells (Zhang et al., 2011). These findings andthose of the present study suggest that the high expres-sion of AQP1 in choroid plexus epithelial cells after 6 hrof hyponatremia triggers the caspase 3-dependent apo-ptotic cascade. Thus, we suggest that increased AQP1 inchoroid plexus epithelium plays a crucial role in casapase

3-dependent apoptosis in addition to the overproductionof CSF following hyponatremia. Furthermore, the inac-tivation of AQP1 after apoptosis may prevent further ap-optotic events despite its high expression after 12 hr ofhyponatremia.

Apart from its possible function in the control ofCSF production, the major physiological role ofNKCC1 may be to promote the movement of Cl–, K1,and Na1 into CSF. Recent findings have demonstratedthat the inhibition of NKCC1 could prevent the occur-rence of AVD and downstream apoptotic events in is-chemia-reperfusion-induced apoptosis in cardiomyocytes(Tanabe et al., 2005). A study conducted using primaryrat cortical neuronal cultures demonstrated that theapplication of Cl– channel blocker is more effective thanthe applications of K1 and Na1 channel blockers againststaurosporine-induced apoptosis, which suggests that Cl–

rather than K1 or Na1 plays a pivotal role in apoptosis(Poulsen et al., 2010). These studies and our results sug-gest that rapid outward Cl– secretion from choroidplexus cells with increased NKCC1 coupled to the reso-lution of osmotic perturbation between plasma and CSFand osmotically obligated water extrusion through non-AQP1-mediated transcellular routes result in choroidplexus cell apoptosis. Thus, apoptosis in choroid plexuscells in response to hyponatremia is, at least in part, con-trolled by the amount of Cl– secretion via NKCC1.Finally, the exact mechanisms of apoptotic unresponsive-ness of choroid plexus epithelial cells after 12 hr of hy-ponatremia must be elucidated, although we believe thatposttranslational nondegradative-inactivation mechanismsexplain in part why epithelial cells respond differently tothe increased expressions of AQP1 and NKCC1.

Selective Necrosis of CA1 Neuronal Cells inResponse to Acute Hyponatremia

Necrosis differs from apoptosis during the earlystages of neuronal injury because loss of plasma mem-brane integrity is one of the initial changes, and it leadsto cell swelling. Although DNA breaks and nuclear pyk-nosis develop during necrosis, they appear after mem-brane damage and cytoplasmic changes and are notaccompanied by peripheral chromatin condensation(Arends et al., 1990). Based on this theoretical back-ground, we confirmed that TUNEL-positive findings inmedial CA1 hippocampus after 6 hr of hyponatremia arepreliminary step in necrotic cell death.

Acute hyponatremia may be considered a metaboliccondition that initiates an increase in ICP by enhancingCSF volume. In general, a greater than 5% increase inCSF volume could induce an increase in ICP (Zander,2009). It has previously been demonstrated that AQP1knockouts exhibit a 56% reduction in ICP vs. wild-typemice under isomolar condition (Oshio et al., 2003).Therefore, necrotic cell death in the hippocampus maybe related to direct compression of hippocampus venulesand capillaries and subsequent reduced cerebral bloodflow resulting from elevated ICP.

6 Kim and Jung

Journal of Neuroscience Research

Alternatively, an increase of CSF in ventricles canenhance the transependymal CSF flow into interstitialspace of the brain, which causes hyposmotic cellularswelling in hippocampus (Tait et al., 2008). Althoughthe exact pathophysiology is not clearly understood, itwas also suggested that cellular swelling triggers massiverelease of excitatory amino acid L-glutamate in neuronalcells but decreases glutamine synthethase activity inastrocytes (Hyzinski-Garcia et al., 2011). These com-bined effects of cellular swelling may be another reasonfor hippocampal cell death in acute hyponatremia.

In this study, apoptotic damage of a large portionof cells in choroid plexus can be expected to causesevere disruption of blood–CSF barrier and normal CSFcomposition. A few reports have examined changes inchoroid plexus cells and/or biochemical alterations inthe CSF following cerebral ischemia (Ikeda et al., 1992;Nagahiro et al., 1994). Therefore, the close anatomicalrelationship between the third ventricle and the hippo-campus increases the possibility that disruption of theblood–CSF barrier accelerates the underlying pathologyof necrotic cell death in the hippocampus.

The hippocampus, specifically the CA1 region, isalso known to be susceptible to elevated ICP as evidenceby various forms of cerebral ischemia and stroke (Dom-browski et al., 2008). Consistent with this finding, inthe present study, cells in the CA1 hippocampal regionappeared to be susceptible to necrotic cell death underhyponatremic conditions, whereas cells in the CA3region were more resistant to apoptotic and/or necroticcell death. Resistance to osmotically induced deforma-tion in CA3 cells could be explained by mechanical re-silience, in part or wholly resulting from the supportprovided to the plasma membrane by the cytoskeleton(Aitken et al., 1998). Thus, these reports and our resultssuggest that different levels of damage in cells or regionsof the hippocampus under hyponatremic conditionsdepend on the intrinsic natures of cells or hippocampalregions even though the increased ICP acts equally. Theexact reason for this heterogeneity is unclear.

In conclusion, the data from the present studydemonstrate that the increased expressions of AQP1 andNKCC1 in choroid plexus induce CSF overproductionand blood–CSF barrier disruption by causing the apo-ptosis of choroidal epithelial cells. These findings suggestthat the deregulations of water and ionic movements inchoroid plexus play pivotal roles in the pathophysiologyof acute hyponatremia.

ACKNOWLEDGMENTS

The authors thank Prof. Jinseo Park for his scien-tific input and Kiyoung Kim for technical assistance.

REFERENCES

Aitken PG, Borgdorff AJ, Juta AJ, Kiehart DP, Somjen GG, Wadman

WJ. 1998. Volume changes induced by osmotic stress in freshly isolated

rat hippocampal neurons. Pflugers Arch 436:991–998.

Arends MJ, Morris RG, Wyllie AH. 1990. Apoptosis. The role of the

endonuclease. Am J Pathol 136:593–608.

Ayus JC, Achinger SG, Arieff A. 2008. Brain cell volume regulation in

hyponatremia: role of sex, age, vasopressin, and hypoxia. Am J Physiol

Renal Physiol 295:619–624.

Di Mattio J, Hochwald GM, Malhan C, Wald A. 1975. Effects of

changes in serum osmolarity on bulk flow of fluid into cerebral ven-

tricles and on brain water content. Pflugers Arch 359:253–264.

Dombrowski SM, Deshpande A, Dingwall C, Leichliter A, Leibson Z,

Luciano MG. 2008. Chronic hydrocephalus-induced hypoxia: increased

expression of VEGFR-21 and blood vessel density in hippocampus.

Neuroscience 152:346–359.

Gagnon KB, England R, Diehl L, Delpire E. 2007. Apoptosis-associated

tyrosine kinase scaffolding of protein phosphatase 1 and SPAK reveals a

novel pathway for Na-K-2C1 cotransporter regulation. Am J Physiol

Cell Physiol 292:1809–1815.

Hasegawa H, Ma T, Skach W, Matthay MA, Verkman AS. 1994. Mo-

lecular cloning of a mercurial-insensitive water channel expressed in

selected water-transporting tissues. J Biol Chem 269:5497–5500.

Hamann S, Herrera-Perez JJ, Zeuthen T, Alvarez-Leefmans FJ. 2010.

Cotransport of water by the Na1-K1-2Cl– cotransporter NKCC1 in

mammalian epithelial cells. J Physiol 588:4089–5101.

Hebert SC, Mount DB, Gamba G. 2004. Molecular physiology of cat-

ion-coupled Cl– cotransport: the SLC12 family. Pflugers Arch 447:580–

593.

Hyzinski-Garcıa MC, Vincent MY, Haskew-Layton RE, Dohare P, Kel-

ler RW Jr, Mongin AA. 2011. Hypo-osmotic swelling modifies gluta-

mate-glutamine cycle in the cerebral cortex and in astrocyte cultures. J

Neurochem 118:140–152.

Ikeda J, Mies G, Nowak TS Jr, Joo F, Klatzo I. 1992. Evidence for

increased calcium influx across the choroids plexus following brief is-

chemia of gerbil brain. Neurosci Lett 142:257–259.

Iwamoto LM, Fujiwara N, Nakamura KT, Wada RK. 2004. Na-K-2Cl

cotransporter inhibition impairs human lung cellular proliferation. Am J

Physiol Lung Cell Mol Physiol 287:510–514.

Jablonski EM, Hughes FM Jr. 2006. The potential role of caveolin-1 in

inhibition of aquaporins during the AVD. Biol Cell 98:33–42.

Jablonski EM, Webb AN, McConnell NA, Riley MC, Hughes FM Jr.

2004. Plasma membrane aquaporin activity can affect the rate of apo-

ptosis but is inhibited after apoptotic volume decrease. Am J Physiol

Cell Physiol 286:975–985.

Jablonski EM, Mattocks MA, Sokolov E, Koniaris LG, Hughes FM Jr,

Fausto N, Pierce RH, McKillop IH. 2007. Decreased aquaporin

expression leads to increased resistance to apoptosis in hepatocellular

carcinoma. Cancer Lett 250:36–46.

Johanson CE. 1993. Tissue barriers: diffusion, bulk flow and volume

transmission of proteins and peptides within the brain. Pharm Biotech-

nol 5:467–486.

Karadshen MF, Byun N, Mount DB, Delpire E. 2004. Localization of

the KCC4 potassium-chloride cotransporter in the nervous system.

Neuroscience 123:381–391.

Liu J, Sun ZH, Shi YL, Wang SY, Wei M. 2010. Inhibitory effect of

siRNA targeting gene Aqp1 on K562 cells. Zhongguo Shi Yan Xue

Ye Xue Za Zhi. 18:40–44.

MacAulay N, Zeuthen T. 2010. Water transport between CNS compart-

ments: contributions of aquaporins and cotransporters. Neuroscience

168:941–956.

Majno G, Joris I. 1995. Apoptosis, oncosis, and necrosis. An overview of

cell death. Am J Pathol 146:3–15.

Melton JE, Nattie EE. 1984. Intracranial volume adjustments and cere-

brospinal fluid pressure in the osmotically swollen rat brain. Am J Phys-

iol 246:533–541.

Milhorat TH, Hammock MK, Fenstermacher JD, Levin VA. 1971. Cere-

brospinal fluid production by the choroid plexus and brain. Science

173:330–332.

AQP1 and NKCC1 in Choroid Plexus 7

Journal of Neuroscience Research

Nagahiro S, Goto S, Korematsu K, Sumi M, Takahashi M, Ushio Y.

1994. Disruption of the blood–cerebrospinal fluid barrier by transient

cerebral ischemia. Brain Res 633:305–311.

Nielsen S, Smith BL, Christensen EI, Agre P. 1993. Distribution of the

aquaporin CHIP in secretory and resorptive epithelia and capillary

endothelia. Proc Natl Acad Sci U S A 90:7275–7279.

Oshio K, Song Y, Verkman AS, Manley GT. 2003. Aquaporin-1 dele-

tion reduces osmotic water permeability and cerebrospinal fluid produc-

tion. Acta Neurochir 86(Suppl):525–528.

Oshio K, Watanabe H, Song Y. 2005. Reduced cerebrospinal fluid pro-

duction and intracranial pressure in mice lacking choroid plexus water

channel aquaporin-1. FASEB 19:76–78.

Pasantes-Morales H, Tuz K. 2006. Volume changes in neurons: hyperex-

citability and neuronal death. Contrib Nephrol 152:221–240.

Payne JA, Forbush B 3rd. 1995. Molecular characterization of the epithe-

lial Na-K-Cl cotransporter isoforms. Curr Opin Cell Biol 7:493–503.

Plotkin MD, Kaplan MR, Peterson LN, Gullans SR, Hebert SC, Delpire

E. 1997. Expression of the Na1-K1-2Cl– cotransporter BSC2 in the

nervous system. Am J Physiol 272:173–183.

Poulsen KA, Andersen EC, Hansen CF, Klausen TK, Hougaard C, Lam-

bert IH, Hoffmann EK. 2010. Deregulation of apoptotic volume

decrease and ionic movements in multidrug-resistant tumor cells: role

of chloride channels. Am J Physiol Cell Physiol 298:14–25.

Spector R, Johanson CE. 1989. The mammalian choroid plexus. Sci Am

261:68–74.

Stahelin BJ, Marti U, Solioz M, Zimmermann H, Reichen J. 1998. False

positive staining in the TUNEL assay to detect apoptosis in liver and

intestine is caused by endogenous nucleases and inhibited by diethyl

pyrocarbonate. Mol Pathol 51:204–208.

Sterns RH, Silver SM. 2006. Brain volume regulation in response to

hypo-osmolality and its correction. Am J Med 119:12–16.

Tait MJ, Saadoun S, Bell BA, Papadopoulos MC. 2008. Water move-

ments in the brain: role of aquaporins. Trends Neurosci 31:37–43.

Tanabe S, Wang X, Takahashi N, Uramoto H, Okada Y. 2005.

HCO3(–)-independent rescue from apoptosis by stilbene derivatives in

rat cardiomyocytes. FEBS Lett 579:517–522.

Vajda Z, Promeneur D, Doczi T, Sulyok E, Frøkiær J, Ottersen OP,

Nielsen S. 2000. Increased aquaporin-4 immunoreactivity in rat brain

in response to systemic hyponatremia. Biochem Biophys Res Commun

270:459–503.

Zander R. 2009. Intracranial pressure and hypotonic infusion solutions.

Anaesthetist 58:405–409.

Zhang L, Li J, Jiang Z, Sun L, Mei X, Yong B, Zhang L. 2011. Inhibi-

tion of aquaporin-1 expression by RNAi protects against aristolochic

acid I-induced apoptosis in human proximal tubular epithelial (HK-2)

cells. Biochem Biophys Res Commun 405:68–73.

Journal of Neuroscience Research

8 Kim and Jung