Expression of Aquaporins in Mouse Choroid Plexus and Ependymal Cells

Wright State University CORE Scholar

Browse all Theses and Dissertations Theses and Dissertations

2015

Pankaj Patyal Wright State University

Follow this and additional works at: https://corescholar.libraries.wright.edu/etd_all

Part of the Pharmacology, Toxicology and Environmental Health Commons

Repository Citation Patyal, Pankaj, "Expression of Aquaporins in Mouse Choroid Plexus and Ependymal Cells" (2015). Browse all Theses and Dissertations. 2037. https://corescholar.libraries.wright.edu/etd_all/2037

This Thesis is brought to you for free and open access by the Theses and Dissertations at CORE Scholar. It has been accepted for inclusion in Browse all Theses and Dissertations by an authorized administrator of CORE Scholar. For more information, please contact [email protected].

EXPRESSION OF AQUAPORINS IN MOUSE CHOROID PLEXUS AND

EPENDYMAL CELLS

A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science

PANKAJ PATYAL

B.Pharm, Chandigarh College of Pharmacy, Punjab Technical University, Jalandhar, 2012

2015

Wright State University

WRIGHT STATE UNIVERSITY

GRADUATE SCHOOL

Date: 08.10.2015

I HEREBY RECOMMEND THAT THE THESIS PREPARED UNDER MY

SUPERVISION BY Pankaj Patyal ENTITLED “Expression of Aquaporins in Mouse

Choroid Plexus and Ependymal cells” BE ACCEPTED IN PARTIAL FULFILLMENT

OF THE REQUIREMENTS FOR THE DEGREE OF Master of Science.

______Francisco Javier Alvarez-Leefmans, M.D., Ph.D. Thesis Director

______Jeffrey B. Travers, M.D., Ph.D. Chair, Department of Pharmacology & Toxicology

Committee on Final Examination

______Francisco Javier Alvarez-Leefmans, M.D., Ph.D.

______David R. Cool, Ph.D.

______J. Ashot Kozak, Ph.D.

______Robert E. W. Fyffe, Ph.D. Vice President for Research and Dean of the Graduate School

Abstract

Patyal, Pankaj, M.S. Department of Pharmacology and Toxicology, Wright State University, 2015. Expression of Aquaporins in mouse choroid plexus and ependymal cells.

The choroid plexuses (CPs) are highly vascularized epithelial structures lying in the brain

ventricles, forming the blood-cerebrospinal fluid (CSF) barrier. CPs are composed of a

monolayer of cuboidal epithelial cells, derived from ependymal cells (EPCs), joined by tight junctions, and surrounding a core of fenestrated capillaries. Choroid plexus epithelial cells (CPECs) secrete most of the CSF through poorly understood mechanisms that involve aquaporins (AQPs) and various solute transport proteins like the Na+-K+-2Cl- cotransporter 1 (NKCC1). It is thought that the concerted action of these membrane proteins directs the transepithelial movement of solutes and water across CPECs. Net water movement across CPECs occurs from the basolateral membrane (blood-facing) through the apical membrane (CSF-facing). However, with the exception of AQP1 in the apical membrane, the molecular pathways for transepithelial water transport are either unknown or their expression is controversial; there are no known AQPs in the basolateral membrane of CPECs to account for the influx of water, and whether AQPs other than

AQP1 are expressed in the apical membrane remains to be elucidated. The expression of

AQP1 in the basolateral membrane is a subject of debate. Previous work from our lab demonstrated strict apical localization of AQP1; its expression in the basolateral membrane of mouse CPECs is undetectable. Thus, AQPs other than AQP1 are potential molecular substrates for putative basolateral water influx. Further, it is not known if

iii

besides AQP1 there are other AQPs in the apical membrane. In previous work from our

group, transcripts for several AQPs were detected in whole choroid plexus tissue using

RT-PCR. Specifically, mRNAs for AQP water channels 0, 1, 2, 4, aquaglyceroporins 3,

7, 9, and “unorthodox” AQP11 were detected, whereas transcripts for AQPs 5, 6, 8, and

12 were undetectable. Because of their relatively high transcript expression the present

work addresses the cellular and subcellular location of AQPs 4, 7 and 9 proteins in

CPECs and EP cells, using immunolabeling with validated antibodies and confocal

immunofluorescence. AQP4 expression in CPECs is debatable. The present results

demonstrate that: 1) AQP 4 is expressed neither in CPECs nor in ependymal cells, but in

radial glial end feet surrounding the basal aspect of EPCs. Thus previous RT-PCR mRNA for AQ4 is likely to represent glial contamination during tissue extraction. 2) AQP 7 expression was found to be restricted to the apical membrane of CPECs and EPCs. 3)

AQP9 was expressed neither in CPECs nor in EPCs. The water pathways for water influx across the CPECs basolateral membrane remain to be elucidated. The functional significance of the present results for CSF secretion and brain extracellular K+ homeostasis is discussed.

Table of Contents

INTRODUCTION ...... 1

Choroid Plexus Epithelia Structure and Function ...... 6

Ependymal Cells ...... 8

Radial Glial Cells ...... 10

Mechanisms of CSF secretion .. ………………………………………………11

Pathways for water permeation across epithelial cells………………………..12

Aquaporins gene protein family…………………………………………….....13

Aquaporin 4…………………………………………………………………16

Localization of AQP4 in other cell types…………..…………….…………19

Aquaporin 7……………………...……………...…………………………..20

Localization of AQP7 in other cell types…………………………..…….....21

Aquaporin 9……………………...……………………………………….....21

Localization of AQP9 in other cell types ………………………..…………23

HYPOTHESIS……...…………………...…………………………………….. .24

Overall Aim……………………………………………………………………24

The Specific Aim of the present work……………………………………...... 25

MATERIALS AND METHODS……………………………………………...... 26

Animals Use………………………………………………………………...26

Perfusion Method…………………………………………………………...26

Antibodies for immunocytochemistry...... 27

Immunolabeling of AQP4……………………………………………….....27

Triple immunolabeling of AQP4, NKCC1 and DAPI…………………...... 29

Quadruplet immunolabeling of AQP4, NKCC1, GFAP and DAPI……...... 31

Immunolabeling of AQP7…………………………………………………31

Immunolabeling of AQP9……………………………………………….....33

Epifluorescence microscopy and confocal imaging……………………...... 34

RESULTS……………………………………………………………………….37

Localization of AQP4 in mouse choroid plexus epithelium and…….…….....

ependymal cells...... 37

Colocalization of AQP4 and NKCC1 in mouse ependyma and capillaries..39

Colozalization of AQP4, NKCC1 and GFAP in ependyma and…………......

subependymal layers………………………………………………………41

Localization of Aquglyceroporin 7 (AQP7) in mouse choroid plexus and.....

ependymal cells………………………………………………………...... 43

Localization of Aquglyceroporin 9 (AQP9) in mouse choroid plexus……...

and ependymal cells……………………………………………………....45

DISCUSSION………………………………………………………………….48

Expression of AQP4……………………………………………………...49

Co-expression of AQP4 and NKCC1………………………………….....51

Expression of AQP7……………………………………………………...54

Expression of AQP9……………………………………………………...55

CONCLUSIONS…………………………………………………………...... 57

REFERENCES…………………………………………………………...58-80

vii

List of Figures

Figure 1. Transcellular pathways for Water flux across choroid plexus epithelial…

cells ...... …..5

Figure 2. Choroid Plexus cerebrospinal fluid (CSF) secretion and flow………….7

Figure 3. Schematic diagram depicting choroid plexus and ependymal cells...... 8

Figure 4. Radial glia cells and their development...... 10

Figure 5. Diagram illustrating the direction of net fluxes of ions and water across .

choroid plexus epithelial cells …………………………………………………..12

Figure 6. Membrane topographical structure of Aquaporin 1 ...... 14

Figure 7.Mammalian Aquaporin Family... .. ……………………………………16

Figure 8. Control immunolabeling in the absence of primary antibody AQP4 in…

mouse kidney cells………………………………………………………………29

Figure 9. Control immunolabeling of NKCC1 antibody and DAPI in choroid……

plexus epithelial cells of letral ventricle(LV) of mouse brain………… ……….30

Figure 10. Control immunolabeling in the absence of primary antibody AQP7

in mouse kidney cells……………………… …………………………………32

Figure 11. Control immunolabeling in the absence of primary antibody AQP7

in mouse kidney cells……………………… …………………………..…….33

viii

Figure 12. Control immunolabeling of AQP9 antibody and DAPI in kidney cells

of mouse kidney cells……..……………………… …………………………….34

Figure 13. Aquaporin 4 water channel protein (AQP 4) immunolabeling in mouse..

lateral brain ventricle structures, kidney proximal tubules and skeletal muscle……

fibers……………………………………………………………………………...38

Figure 14. Confocal images of triple immunolabeling of AQP4 (red), NKCC1…...

(green) and DAPI (blue) in choroid plexus epithelial cells (CPECs) and………….

ependyma (EP) of lateral ventricle (LV) of mouse brain………………………..40

Figure 15. Confocal images of quadruplet immunolabeling of NKCC1, AQP4,….

GFAP and DAPI in choroid plexus epithelial cells (CPECs) and ependyma (EP) ...

of lateral ventricles (LV) of mouse brain………………………………………..42

Figure 16. Aquaporin 7 water channel protein (AQP 7) immunolabeling in mouse. lateral brain ventricle structures, kidney proximal tubules and skeletal muscle…… fibers...... 44

Figure 17. Aquaporin 9 water channel protein (AQP 9) immunolabeling in mouse..

brain thalamus, the auditory neocortex and liver bile duct……………………….46

Figure 18. Diagram illustrating the distribution of AQP4 and NKCC1 in the ……..

endings of radial glia and in the ependy……………………………………….…54

List of Tables

Table1. Antibodies used in this study …………………………………….36

Acknowledgements

I acknowledge my profound and sincere thanks to my advisor, Dr. Francisco J.

Alvarez-Leefmans. I thank him for accepting me into his laboratory and assigning me a

thesis project that turned out very exciting. It has been a very gratifying experience to

work with him and learn from him. I cannot be grateful enough for all the support and

advice that he has given me, as well as the numerous hours that he has invested in my

formation as a fellow scientist. He has been teaching me all about self-discipline in

laboratory work and in written scientific communication.

I am indebted to my committee members Dr. David R. Cool and Dr. J Ashot

Kozak. They have provided, with kindness, their insight and suggestions, which are

precious to me.

I would like to thank Ms. Jeannine M. Crum for her help in many of the technical

aspects involved in the present work. Likewise, I wish to thank Mr. David Cha for his

support and help in some aspects of this work.

The results included in the present study were obtained in collaboration with Dr.

F.J. Alvarez-Leefmans (thesis director) and therefore are not my sole intellectual

property.

Dedication

To my beloved- Guruji, parents, Dr. Kamlesh and Mrs. Pushpa; and younger sister Neha

Thank you for their everlasting love and support

To all the best people, I met in my life.

xii

INTRODUCTION

The choroid plexuses (CPs) of the brain ventricles secrete cerebrospinal fluid

(CSF) at a rate of about 600 milliliters per day. This high fluid transport rate is similar to

that of the renal proximal tubules (Damkier et al., 2013). The CSF is constantly

reabsorbed, so that only ~150 ml are present at any one time in the entire CNS, thus CSF

is replaced four times per day. The CPs are highly vascularized epithelial structures that

form the blood-cerebrospinal fluid barrier (BCSFB). The mechanisms of fluid secretion

by the CPs are not understood. Moreover, the pathways through which water moves

across the plasma membranes of choroid plexus epithelial cells (CPECs) are largely unknown. Given the relatively high rate of fluid secretion by the CPECs, water flow must occur through specialized plasma membrane proteins, such as aquaporins (AQPs), rather than through poorly water-permeable lipid bilayers (Friedman, 2008). The objective of this study was to analyze the protein expression of three of these aquaporins that are thought to exist in the CP, to determine if they might play a role in water flux across the

BCSFB.

Like other epithelial cells, CPECs are polarized; having an apical membrane

(CSF-facing), and a basolateral membrane (blood-facing). Among many unanswered questions in CP function, are the mechanisms underlying transepithelial water transport.

It is well established that aquaporin 1 (AQP1) is expressed in the apical membrane of

CPECs (Nielsen et al., 1993). To date, there are no known AQPs in the basolateral

membrane and apical AQP1 does not fully explain the rate of cerebrospinal fluid

secretion. Moreover, in AQP1 KO mice CSF secretion rate is reduced by only 25%

(Oshio et al., 2003, 2005). Thus, AQPs other than AQP1 are potential molecular

substrates for putative basolateral water influx. Further, it is not known if besides AQP1

there are other AQPs in the apical membrane. The specific objective of the present study

was to investigate the cellular and subcellular expression of AQPs 4, 7 and 9 proteins in

CPECs and ependymal cells (EPCs), the cells that give rise to CPECs.

CPs are composed of a monolayer of cuboidal epithelial cells derived from

ependymal cells (EPCs), joined together near their apical lateral surface by tight junctions

(Prescott and Brightman, 2005). The CPs surround a core of capillaries that are

fenestrated and do not have tight junctions, unlike those forming the blood-brain barrier

(BBB) that are not fenestrated and do have tight junctions. It is these features of CP capillaries that allow water and small solutes to move freely through the endothelia between the blood and basolateral membrane of the CPECs (Prescott and Brightman,

2005).

It is known that net transcellular water flux across CPECs occur from the basolateral membrane to the apical membrane i.e. from blood to the CSF, however the proteins and the mechanisms that carry out fluid secretion are largely unknown (Damkier

et al., 2010, 2013). One of the main puzzles is that until now, putative AQPs have not

been demonstrated in the basolateral membrane (Damkier et al., 2013).

Transepithelial water transport, in general, can occur through transcellular and paracellular routes (Friedman, 2008). The paracellular transport route refers to the transfer of solute and fluid through the intercellular spaces between the cells, via tight junctions. Whereas transcellular transport implies that solute can flow through the cells

passing through both apical and basolateral membranes (Friedman, 2008; Damkier et al.,

2013). In CPECs it is thought that water movement occurs mainly, if not exclusively,

through transcellular pathways, but so far, possible paracellular pathways have not been

investigated (Damkier et al., 2013). There is evidence for paracellular water transport

across MDCK C7 epithelial monolayer cultures stably transfected with claudin-2, a tight

junction protein (Rosenthal et al., 2010). Tight junctions (TJs) are composed of different

transmembrane proteins, including occludin and members of the claudin family. Claudins

are small (20–27 kDa) transmembrane proteins, which support the epithelial barrier

function and also form paracellular channels (Anderson and Van Itallie, 1995). Although

native CP TJs expresses claudin-2, there is no proof for a paracellular pathway for water

transport in CP epithelia. Moreover, although CP is often considered a “leaky”

epithelium, there is no solid proof for this assertion (Damkier et al., 2013). On the

contrary, given its role in the formation of the BCSFB, the CP is expected to be a “tight”

epithelium, i.e. it must have a transepithelial resistance well above 100 ohm-cm2

(Friedman, 2008). However, no reliable measurements of transepithelial resistance exist

for mammalian CP, except for those in cultured monolayers, which range between 150 and 1,500 ohm-cm2 depending on presence or absence of serum (Hakvoort et al., 1998;

Angelow and Galla, 2005).

Since water flows across CPECs from the basolateral through the apical

membrane, there must be one or more water pathways in both apical and basolateral

membranes. There are three possible pathways for water permeation across cell

membranes: 1) AQPs (Papadopoulos and Verkman, 2013), 2) cotransporters (Hamann et

al., 2005, 2010; Zeuthen and Macaulay, 2012), and 3) the lipid bilayer (Finkelstein, 1976;

Krylov et al., 2001). There is evidence for water transport through cation-coupled

chloride cotransporters like NKCC1 and K+-Cl- (KCCs) (Zeuthen, 1994; Hamann et al.,

2005). However, in the CP these cotransporters mediate vectorial water transport in the

direction opposite to that expected for net trans-epithelial water movement (Fig.1). There is substantial evidence for AQP1 expression in the apical membrane of CPECs (Nielsen et al., 1993; Damkier et al., 2013). However, it is debatable whether AQP1 is also expressed in the basolateral membrane (Johansson et al., 2005; Praetorius and Nielsen,

2006; Papadopoulos and Verkman, 2013). Immunolabeling results from the Alvarez-

Leefmans laboratory, using confocal immunofluorescence, demonstrate that AQP1 is not expressed in the basolateral membrane (Cha et al., 2014), an observation that is in agreement with the original immunogold electron microcopy results of Agre’s group

(Nielsen et al., 1993). Interestingly, CSF production is reduced only by 25% in AQP1-/- mice (Oshio et al., 2003, 2005). This suggests that AQP1 is not the only apical water pathway underlying CSF secretion, and the water pathways for water influx across basolateral membranes of CPECs are unknown. Likewise, whether AQPs other than

AQP1 are expressed in the apical membrane of CPECs is unknown (Fig. 1).

Figure1. Transcellular pathways for water transport across choroid plexus epithelial cells. Unlike all secretory epithelial cells, CPECs co-express NKCC1 (Na+, K+, 2Cl- cotransporter 1) and Na+/K+ ATPase in the apical membrane, together with AQP1. The pathways for water influx in the basolateral membrane are unknown (AQPx in the basolateral membrane). It is not understood how transcellular water movement occurs through and from the basolateral through the apical side. Further, there might be other AQPs beside AQP1 on the apical membrane (AQPx). (Courtesy of Dr. F. J. Alvarez-Leefmans, 2015).

Previous research from the Alvarez-Leefmans lab using RT-PCR demonstrated the existence of transcripts for several AQPs in whole choroid plexus tissue. Specifically, mRNAs for AQP water channels 0, 1, 2, 4, aqua-glyceroporins 3, 7, 9, and “unorthodox”

AQP11 were detected. Transcripts for AQPs 5, 6, 8, and 12 were undetectable (Cha et al.,

2014).

The global aim of this project is to search for the cellular and subcellular expression of a subset of AQP proteins whose transcripts were detected by RT-PCR (Cha

et al., 2014). The present work is focused on the cellular and subcellular localization of

AQPs 4, 7 and 9 using immunolabeling methods with validated antibodies and confocal

immunofluorescence microscopy.

Choroid Plexus Epithelium Structure and Function

As mentioned above, the choroid plexuses are branched and highly vascularized

epithelial structures derived from the ependymal cells lining the brain ventricles

(Damkier et al., 2013, Barry et al., 2014). There are four choroid plexuses inside the ventricular system of the brain; one in each of the two lateral ventricles, one in the third,

and one in the fourth ventricle (Fig.2A). The choroid plexuses are composed of a

capillary enveloped by a layer of differentiated ependymal epithelium (Fig. 3). This

monolayer of epithelial cells that are joined by tight junctions, make up the BCSFB

(Smith et al., 2004). In humans, the CP secretes about 600 ml of CSF in 24 hours, a rate

that is similar to that of renal proximal tubules (Damkier et al., 2013). Further, the blood

flow in CP vasculature is about 5 times greater than that of kidney (Damkier et al., 2013).

Since the total volume of CSF is about 150 ml, this fluid is replaced approximately 4

times a day. In addition to CSF production, the CP acts as a filtration system, removing

metabolic waste, foreign substances, and neurotransmitters from CSF. The kidneys

remove waste material from blood to urine; whereas, the CPECs remove waste materials

from CSF to blood. Therefore, the CP has been considered as “the kidney of brain”

(Spector and Johanson, 1989). In sum, the CP has a crucial role in maintaining the

composition of the brain extracellular environment required for proper brain function.

The CSF secreted in the lateral ventricles flows through the Foramina of Monro into the

third ventricle and then through the aqueduct of Sylvius into the fourth ventricle. CSF leaves the ventricular system via the Foramina of Luschka and Magendie into the subarachnoid space (Damkier et al., 2013) (Fig.2).

Figure 2. Choroid Plexus cerebrospinal fluid (CSF) secretion and flow. A) The diagram of human brain in which arrows indicate the flow direction of CSF. CP in the lateral ventricle secretes the CSF which flows through the foramen of Monro into third ventricle. From third ventricle, CSF flows into the fourth ventricle via the aqueduct of Sylvius. From there it flows into the central canal of the spinal cord or into the cisterns of the subarachnoid space via three small foramina: the central foramen of Magendie and the two lateral foramina of Luschka. B) The amplified diagram showing how the secretion of CSF from CPECs diffuses through ependyma. CSF reabsorbs into the venous blood through the arachnoid villi in the superior sagittal sinus. (A reproduced from, Grapp et al., 2013, Choroid plexus transcytosis and exosome shuttling deliver folate into brain parenchyma, Nature Communications, 4, 2123. B reproduced from, Rosenberg, G.A. Brain Fluids and Metabolism. Oxford University Press, 1990)

Figure 3. Schematic diagram depicting choroid plexus and ependymmal cells. The choroid plexuses are formed by fenestrated capillaries surrounded by a single layer of choroid plexus epithelial cells. These cuboidal-shaped cells are modified ependymal cells joined together by tight junctions. The choroid pllexuses float in the ventricular cavities surrounded by the CSF they secrete. (Modified from: http://www.daviddarling.info/encyclopedia/ C/choroid_plexus.html)

Ependymal Cells Ependymal cells are ciliated epithelial cells derived from the neuroepithelium

(Jimenez et al., 2014). They form the lining of the brain ventricles and the central canal of the spinal cord. Ependymal cells are located at the interface between brain parenchyma and the CSF-filled ventricles in the central nervous system (CNS) (Jimenez et al., 2014).

During CNS development, the neuroepithelium gives rise to a subset of radial glial cells that differentiate into ependymal cells (Spassky et al., 2005; Jimenez et al., 2014; Barry et al., 2014). Mature ependymal cells are cuboidal in shape and are polarized. In their apical membrane they display motile cilia and microvilli. The cilia beating of these cells are

crucial for CSF circulation (Jimenez et al., 2014). Ependymal cells are loosely connected

by intercellular adhesion sites known as desmosomes, or adherens junctions thereby

forming an epithelial sheet that constitutes the lining of the ventricles. Unlike CPECs,

mammalian ependymal cells do not have tight junctions (Peters and Swan, 1979; Jimenez

et al., 2014). This may explain why ependymal cells do not form a diffusion barrier

between brain parenchyma and CSF and thus, brain interstitial fluid composition

determines and is determined by CSF composition as illustrated in Fig. 2B. Specifically,

the ependyma is permeable to ions, water and even proteins (Jimenez et al., 2014).

Indeed, several studies demonstrate that the intercellular spaces between ependymal cells

allow the passage of water, ions and proteins. For instance, the ependymal intercellular

spaces are permeable to lanthanum, peptides and proteins such as horse radish peroxidase

and ferritin that have molecular mass of 43 and 560 kda respectively (Brightman, 1965;

Cifuentes et al., 1992).

Diffusion of CSF usually occurs from brain ventricles into the brain parenchyma

through these loosely bound junctions between the ependymal cells. (Hladky and

Barrand, 2014). In contrast, CPECs are modified ependymal cells joined by tight

junctions which prevents “leakage” of fluid from blood vessels into CSF. Early studies

suggested a role of ventricular ependymocytes in the secretion of CSF (Pollay and Curl,

1967). Furthermore, it was reported that CSF production was reduced but not completely

eliminated by choroid plexectomy (Milhorat et al., 1971). However, this is controversial

inasmuch as the there are no mechanistic models of how CSF could be secreted by

ependymal cells, and there is no proof that choroid plexotomy is complete. Thus, it is

unlikely that ependymocytes secrete CSF.

Radial Glial Cells

Radial glia are elongated bipolar-shaped cells which differentiate from neuroepithelial cells during CNS development. In the CNS these cells extend their processes from the ventricular surface to the pial surface (Götz, 2013). Radial glial cells are multifunctional; they can serve as scaffolds for radial migration of neurons, but a subpopulation of radial glial cells differentiate into neurons (Lui et al., 2011). Radial glia in the cortex contributes to neurogenesis through neuronal precursor cells (nIPC). The radial glia of cortex differentiates into astrocytes and oligodendrocytes by gliogenesis.

Later, some radial glia also differentiate into ependymal cells which line the ventricles of the adult brain and these ependymal cells give rise to choroid plexus epithelial cells

(Barry et al., 2014) (Fig. 4).

See Figure 4 legend on next page.

Figure 4. Radial glial cells and their development. Neuropepithelial (NEP) cells proliferate and give rise to neuroblasts, which differentiate into neurons. NEP cells also give rise to radial glia which can contribute to neurogenesis directly or via neuronal precursor cells (nIPC), and also differentiate into astrocytes and oligodendrocytes. Radial glial can also differentiate into ependymal cells, which in turn are modified to form choroid plexus epithelial cells. (Reproduced from Barry, D. S., Pakan, J. M., & McDermott, K. W. 2014, Barry et al., 2014)

Mechanisms of CSF secretion

CSF is mostly formed inside the cerebral ventricular system. Choroid plexuses

epithelial cells are known to be the main if not the only source of CSF production (Dandy

1919; Segal, 1993; Johanson et al., 2008; Chikly and Quaghebeur, 2013). Extrachoroidal

CSF sources, such as the ependymal cells (Pollay and Curl, 1967; Milhorat et al., 1971) or the BBB capillaries (Abbott, 2004) have also been proposed but this is debatable. The endothelium of the choroid plexus capillaries is fenestrated, and CSF formation is determined by the passage of a plasma ultrafiltrate through the endothelium, which is facilitated by hydrostatic pressure. The choroidal epithelium secretes fluid, into the ventricles. The passage of solute and water through the choroidal epithelium is an active metabolic process which transforms the ultrafiltrate into a secretion (cerebrospinal fluid)

(Brown et al., 2004). Fluid secretion in choroid plexus epithelia has been found to be dependent on the unidirectional transport of ions, which creates osmotic gradient for the movement of water (Fig. 5). There are no known pathways for water transport across

CPECs through basolateral membrane. The basic understanding of this process is

incomplete and still speculative (Damkier et al., 2013). Thus, one of the goals of this

project was to provide insight into the mechanisms and pathways underlying net water

flux from blood into the CSF.

Figure 5. Diagram illustrating the direction of net flux of ions and water across choroid plexus + - - epithelial cells. The net flux of Na , Cl , HCO3 , and water are in the direction from blood to cerebrospinal fluid (CSF). outwards into the lumen, that is coming from the fenestrated capillaries; and the net flux of K+ is inwards back into the circulatory system (Courtesy of Dr. Alvarez-Leefmans, 2015).

Pathways for water permeation across epithelial cells The net transepithelial water transport can occur through transcellular and

paracellular routes. Accordingly, the net water movement occurs from the basolateral

through the apical membrane of CPECs. As already mentioned, water permeation across

cell membranes occur through three pathways: aquaporin channels (AQPs), some

cotransporters, and to a lesser extent through the membrane lipid bilayer. The intrinsic

water permeability of lipid bilayers is too low to explain the high volumes of CSF

secretion (500-600 ml/24 hrs in humans).Thus, water movement explaining this large

fluid secretion rates must occur through aquaporins. Cotransporters cannot explain water

flux across CPECs because they transport ions (and possibly water) in the direction

opposite to that expected for net water flux (Damkier et al., 2013). Thus, the most

plausible explanation for such a large amount and rate of fluid secretion is that channel

proteins with the features of aquaporins must be involved.

The role of paracellular pathways in transepithelial water movement is debatable

(Fischbarg, 2010). A possible role of claudin-2 as a paracellular channel that mediates water transport across the tight junctions in “leaky” epithelia has been suggested (Gunzel and Yu, 2013). There is evidence for a paracellular water pathway across claudin-2 tight junctions’ as discussed above (Rosenthal et al., 2010).

Aquaporin gene-protein family

Aquaporins are a family of integral membrane proteins some of which transport water across cell membranes in response to osmotic gradients (Papadopoulos et al. 2013).

Aquaporins and Aquaglyceroporins are monomers of molecular weight of ~26-34 kDa that contain six membrane-spanning helical segments and two shorter helical segments that do not span the entire membrane (Beitz, 2009; Thomas et al., 2002). These monomers consist of six transmembrane helical segments and two short helical segments

that surround cytoplasmic and extracellular vestibules connected by a narrow aqueous

pore (Walz et al., 1997; Verkman, 2013; Beitz, 2009) (Fig. 6). AQPs usually form

tetramers with each monomer functioning as an independent pore.

Figure 6. Membrane topographical structure of Aquaporin 1. This structure consists of eight membrane-embedded helical segments (H1–H8) between the N- and C-terminus. (Reproduced from: Verkman et al., 2013)

The search for a “water channel” accounting for high water flux across epithelial

cells, that could not be explained by simple diffusion though lipid bilayers, culminated

with the discovery in the early 90s of CHP 28 (currently known as AQP1), by Peter Agre

and his coworkers (Preston et al., 1992; Nielsen et al., 1993). CHIP 28 (Channel-Forming

Integral Protein of 28 kDa) was isolated from water permeable segments of the nephron

and red blood cells. It was found to be an integral membrane protein with a molecular

mass of ~28 kDa. The discovery of CHIP 28 ended long-standing controversies about the

existence of membrane water channel proteins.

To date, thirteen mammalian aquaporins (0-12) have been identified and partially

characterized from the molecular and functional perspectives (Verkman, 2011). These

proteins can be divided into three categories, depending on their functional properties

such as water permeation and selectivity: i) those selectively permeable to water

(aquaporins); ii) those permeated by water and few other polar molecules like glycerol

(aquaglyceroporins), and iii) those whose functions are unknown and are currently being elucidated, here referred to as unorthodox (Xu et al., 2010) (Fig.7). The mammalian

Aquaporins consist of AQP0, AQP1, AQP2, AQP4 and AQP5 function primarily as

water-selective channels. Cells expressing AQP water channels in their plasma

membrane have ~5- to 50-fold higher osmotic water permeability than membranes which

do not have AQPs (Verkman and Mitra, 2000). AQP3, AQP7, AQP10 and AQP9 are the

aquaglyceroporins; they transport water, glycerol, and, in the case of AQP9 they also

transport other small polar solutes, including amino acids, sugars and arsenite (Verkman,

2011). AQP6, AQP8, AQP11 and AQP12 are classified as “unorthodox” aquaporins, as

their functional characteristics are little known or unknown (Xu et al., 2010).

Furthermore AQPs have also been reported to transport other small molecules and gases,

including carbon dioxide, ammonia, nitric oxide and hydrogen peroxide (Musa-Aziz et

al., 2009; Miller et al., 2010). The present work focuses on AQP 4, 7 and 9, thus the

following sections deal specifically with these AQPs.

Figure 7. Mammalian Aquaporin Family. The aquaporin gene family leads to proteins that are generally permeated only by water (AQP0, AQP1, AQP2, AQP4, AQP, 5). The aquaglyceroporins are permeated by water and small solutes such as glycerol (AQP7, AQP3, AQP9, AQP10). The unorthodox (AQP11, AQP12, AQP8, AQP6) are members with little known functions. AQP10 is a pseudogene in mice. (Modified from: Xu et al; 2010)

Aquaporin 4

Aquaporin 4 (AQP4) is a transmembrane protein which functions as molecular water channel. It is also known as the “mercurial-insensitive water channel” (MIWC)

(Yang and Verkman, 1996). Its cloning from lung and brain was described in 1994 by two independent groups. (Hasegawa et al., 1994; Jung et al., 1994).

We chose to investigate the expression of AQP4 in choroid plexus epithelium and in

ependymal cells because of its potential functional implications in CSF secretion and

brain water homeostasis. Another strong reason that motivated the search for AQP4

expression in CP is its abundant mRNA expression, as evidenced from RT-PCR work

from the lab, on mouse whole choroid plexus (Cha et al, 2014), and from reports in the

literature showing its expression in CPECs using in situ hybridization (Venero et al.,

1999), immunolabeling (Mobasheri et al., 2007; Speake et al., 2003), and immunoblots

(Nazari et al., 2014). Specifically, the in situ hybridization results of Venero’s group

(Venero et al., 1999) reported a strong hybridization signal of AQP4 mRNA in the choroid plexus of the lateral ventricles. However, the expression of AQP4 in CPECs has been studied by other investigators showing that the AQP4 mRNA is not expressed in

CPECs. Accordingly, using in situ hybridization two independent groups were unable to

detect AQP4 hybridization signal in CP (Hasegawa et al., 1994; Jung et al., 1994). Using

immunoperoxidase with affinity purified antibodies (Frigeri et al., 1995) demonstrated

AQP4 (MIWC) immunoreactivity in rat astrocytes and ependyma, but CPECs were not

immunoreactive. Similarly, Soren Nielsen’s group using immunogold EM was unable to

demonstrate AQP4 immunoreactivity in CPECs (Nielsen et al., 1997).

The immunolabeling studies purporting the expression of AQP4 in CPECs

deserve further comment. None of the three studies reporting AQP4 expression in CPECs

(Mobasheri et al., 2007; Speake et al., 2003; Nazari et al., 2014) were done with

antibodies validated against AQP4 -/- animals. The studies by Speake (Speake et al.,

2003) and Mobasheri (Mobasheri et al., 2007) showed a diffuse cytoplasmic pattern of

putative AQP4 immunoreactivity, whereas that of Nazari claimed that AQP 4 was

“densely packed in the cytoplasm” of CPECs with no expression in the membrane

(Nazari et al., 2014). However, the latter authors did not show these results. From the

above data it can be concluded that the expression of AQP4 in CPECs cells is

controversial. The present study specifically addresses this issue to solve this

controversy.

As for the case of ependymal cells, using in situ hybridization, AQP4 mRNA

expression has been reported by Hasegawa and Jung (Hasegawa et al., 1994; Jung et al.,

1994) in the ependymal cells lining various brain areas. However, it is not clear from the

images published by these authors to determine if the observed hybridization signal is in the ependymal cells or in surrounding cellular elements such as the multiple glial ending process that exist in between and in the basal region of ependymal cells. A similar conclusion can be reached by examining the first paper reporting AQP4 expression in the ependyma. These authors found that AQP4 (then named MIWC) was “strongly expressed in the ependymal layer lining the aqueductual system” (Frigeri et al 1995). However, since they did not use glial markers it is not possible to ascertain whether the immunoreactivity was localized to the ependymal cells or to the surrounding glial membranes. Immunolabeling studies using peroxidase immunoreactivity and immunogold-EM showed that subpopulations of ependymal cells, specifically those covering the subfornical organs and other osmosensory areas, expressed AQP4 in the

basolateral membrane of ependymocytes (Nielsen et al., 1997). The authors cautiously

stated that they found “moderate densities of gold particles in the basolateral membranes

of the ependymal cells covering the subfornical organ,” whereas the apical membranes

were free of labeling. In contrast they found robust expression of AQP4 in glial

membranes that are in direct contact with capillaries and the pia mater. Other studies

(Roales-Bujan et al., 2012; Jimenez et al., 2014) using immunolabeling with non- validated AQP4 antibodies claim that AQP4 is expressed in the apical and basolateral membranes of mouse lateral ventricle ependymal cells. Thus, whether AQP4 is expressed in ependymal cells other than those in the subfornical organ is debatable. The issue of whether AQP4 is expressed in ependymocytes and/or surrounding glial membranes is specifically addressed in the present study using multiple immunolabeling with glial

markers.

Localization of AQP4 in other cell types

Unlike the debatable expression of AQP4 in CPECs and EPCs discussed above,

there is overwhelming evidence for AQP4 expression in astrocytes (Frigeri et al., 1995;

Nielsen et al., 1997; Rash et al., 1998; Nagelhus et al., 2004), using antibodies validated

against KO tissues (Katada et al., 2014). Expression of AQP4 is very prominent in

astroglial endfeet, particularly at perivascular glial processes, subpial membranes and the

ventricular surface (Nielsen et al., 1997; Nagelhus et al., 2004).

In extracerebral tissues, AQP 4 is expressed in kidney, skeletal muscle, salivary

and lacrimal glands (Frigeri et al., 1995). In rat kidney AQP4 was found mainly in the

proximal tubules of the inner medullary collecting ducts (IMCD). Immunoelectron

microscopy demonstrated AQP4 labeling of the basolateral membrane of IMCD cells,

with relatively little labeling of intracellular vesicles (Terris et al., 1995). Using AQP4

antibodies validated against AQP4-/- tissues, Van Hoek (Van Hoek et al., 2000)

demonstrated AQP4 is expressed in basolateral membranes of proximal tubules S3

segments in mouse kidney. Also, AQP4 is highly expressed in the sarcolemma of skeletal muscle fibers (Frigeri et al., 1995; Crosbie et al., 2002; yang et al., 2000). These tissues

(skeletal muscle and kidney) were used in present work as positive controls.

Aquaporin 7

Aquaporin 7 (AQP7), another member of the AQP gene family, is a transmembrane protein permeable to water, glycerol and urea and therefore it is an aquaglyceroporin. AQP7 was first cloned from rat testis by Ishibashi’s group (Ishibashi et al., 1997). This 26-KDa protein was functionally characterized by the same group using heterologous expression in Xenopus oocytes. Osmotic water permeability across oocytes membrane was increased by 10-fold with respect to uninjected oocytes. Another significant feature is that the osmotic water flux through AQP7 was not inhibited by 0.3 mM mercury chloride.

We decided to investigate the protein expression of AQP7 in CPECs and EPCs because of its abundant mRNA expression in whole CP tissue (Cha et al., 2014), and because little is known about the expression of this protein in CPECs and EPCs where it may play a role in water transport together with AQP1. There is only one published study showing diffuse AQP 7 immunoreactivity in both CPECs and EPCs, but its subcellular localization i.e. apical or basolateral membrane on CPECs and/or EPCs was not investigated (Shin et al., 2006). In their study, Shin’s group used a well characterized

polyclonal antibody (sc-28625, SC), which is the same used in the present work and

which has been validated in several studies (Laforenza et al., 2010, 2013; Goubau et al.,

2013; Hermo et al., 2008; Aure et al., 2014). AQP7 is expressed in adipocytes

(Fruhbeck, 2005, 2006; Maeda et al., 2008, 2009) and in endothelial cells within human adipose tissue. (Rodriguez et al., 2011; Laforenza et al., 2013). AQP7 has significant influence on glycerol metabolism by facilitating glycerol transport between adipocytes and blood (Lebeck, 2014). AQP7 KO mice are obese (Hara-Chikuma et al., 2005). The lack of AQP7 results in an increased accumulation of glycerol within adipocytes leading to their hypertrophy (Hibuse et al., 2005).

Localization of AQP7 in other cell types

AQP7 has been found in rat testis by real-time PCR (Sakai et al., 2014). The expression of AQP7 has been studied in the testis and sperm. It is highly expressed in sperm (Sohara et al., 2007). It is shown in more detail in the rat testis by an immunohistochemical method (Suzuki-Toyota et al., 1999). Nejsum and Nielsen group

(Nejsum et al., 2000) demonstrated the expression of AQP7 protein in rat and mouse kidney where it was found to be localized at the apical membrane of the proximal tubules. Also, AQP7 is highly expressed in mouse sarcolemma of skeletal muscle fibers and it has been extensively investigated at mRNA and protein levels (Wakayama et al.

2004). These tissues (skeletal muscle and kidney) were used as positive controls in present study.

Aquaporin 9

Aquaporin 9 (AQP9), another member of the aquaglyceroporin subfamily is a transmembrane protein that allows permeation of water, glycerol and urea. It was first cloned and identified from human leukocytes and liver in 1998 by Ishibashi’s group

(Ishibashi et al., 1998). This 31.4-KDa protein was functionally characterized by the

same group using heterologous expression in Xenopus oocytes. Osmotic water permeability of oocytes membrane was increased by 7-fold with respect to uninjected oocytes, and it was inhibited by 0.3 mM mercury chloride. We investigated the protein

expression of AQP9 in CPECs and EPCs because of its abundant mRNA expression in

whole CP tissue (Cha et al., 2014), and because its possible expression in choroid plexus has not been studied. Further, AQP9 expression in ependymal cells and astrocytes is controversial (see below).

Using double immunolabeling with anti-GFAP, Badaut’s group (Badaut et al.,

2004) found expression of AQP9 in astrocytes and glia limitans. AQP9 labeling was clear in astrocytic processes and their cell bodies. Badaut’s group used a well characterized rabbit polyclonal antibody (AQP91-A, alpha), which is the same used in the present

study. This antibody has been validated against AQP9 KO mouse tissues (Mylonakou et al., 2009).

AQP9 transports glycerol and highly expressed in liver cells, where it serves as the main glycerol channel. AQP9 abundance changes in parallel with alterations in the glycerol permeability of hepatocytes (Calamita et al., 2012). It has been reported that

AQP9 plays a pivotal role in supplying glycerol for gluconeogenesis (Jelen et al., 2011).

There are conflicting reports about AQP9 expression in ciliated ependymal cells and astrocytes (Albertini and Bianchi, 2010). AQP9 immunolabeling was found to be associated with cells, surrounding the cerebral ventricles, including ependymal cells and tanycytes in the hypothalamus (Elkjaer et al., 2000). In the same study they were unable to detect AQP9 immunosignal in astrocytes surrounding the ventricles (Elkjaer et al.,

2000). On the other hand, Badaut’s group reported astrocytic AQP9 expression in the mouse brain (Badaut et al., 2001), but they did not find ependymal immunostaining.

AQP9 was found to be expressed in different subsets of glial cells in the adult forebrain periventricular region and in the ependymal and subependymal layer co-expressed with

GFAP (Cavazzin et al., 2006). Clearly, the expression of AQP9 in ependymal cells and astrocytes is debatable (Albertini and Bianchi, 2010). The issue of whether AQP9 is expressed in ependymocytes and/or surrounding (subependymal) glial membranes is specifically addressed in the present study using multiple immunolabeling with glial markers.

Localization of AQP9 in other cell types

Using real time PCR, AQP9 mRNA was found in rat liver, brain and testis, but it was absent in kidney (Sakai et al., 2014). AQP9 mRNA was also detected in mouse liver cells (Kuriyama et al., 2002). AQP9 protein expression in mouse liver bile ducts and testis were ascertained by immunolabeling (Nicchia et al., 2002). Badaut (Badaut at al.,

2004) also reported the presence of AQP9-IR in mouse liver, whereas no AQP9-IR was found in kidney cells, using the same KO validated antibody used in the present study.

HYPOTHESES

The high rate of CSF secretion via CPECs cannot be easily explained by the

presence of a single apical aquaporin such as AQP1. This assertion is supported by

experiments showing that CSF secretion rate is reduced by only 25% in AQP1 KO mouse

(Oshio et al., 2003). If transcellular water transport underlies CSF secretion, there must be other AQPs in CPECs besides apical AQP1.

We hypothesize that there are AQPs other than AQP1 localized to the apical membrane and to the basolateral membrane of CPECs explaining the water flux across

CP. We also hypothesize that ependymal cells, which are the precursors of CPECs, already express some of the proteins of CPECs. These hypotheses began to be tested in previous work in the lab using RT-PCR with primers designed to amplify transcripts for all AQPs. Gene profiling of all known mRNAs of the AQP gene family in mouse CP showed that besides AQP1, whole CP tissue expresses transcripts for AQPs 0, 2, 3, 4, 7, 9

& 11.

Overall Aim

To investigate the cellular and subcellular localization of AQP proteins in CPECs

and ependyma using immunolabeling with validated antibodies, and confocal

immunofluorescence.

The Specific Aim of the present work

To determine the cellular and subcellular localization of AQP4, AQP7, and AQP9 in CPECs and EPCs.

MATERIALS AND METHODS

Animal Use

All animal experiments were done following the National Institutes of Health

(NIH) guidelines and were carried out with a protocol approved by the Wright State

University (WSU) Laboratory Animal Care and Use Committee (LACUC). The immunolabeling analysis was performed in tissues obtained from two mouse lines named as follows: NKCC1|PV-Cre|GFP| and NKCC1 KO and wild type mice (Mus musculus) on mixed background (C57BL/129vJ) and (NKCC1 KO 129jv/BS line breeders) respectively. Five wild type (WT) mice of various postnatal ages were used: P 12

(NKCC1|PV-Cre|GFP|), P 19 (NKCC1 KO 129jv/BS), P 20 (NKCC1|PV-Cre|GFP|), and

P 30(2 mice, one from each line). No differences were found in AQP4, 7 and 9 expressions between WTs irrespective of their background and postnatal age. In addition we used two P29 NKCC1 knockouts (NKCC1|PV-Cre|GFP|) to validate anti-NKCC1 antibody. Surgical procedures like non-survival surgery and perfusion were performed in the Microscope Core Facility (MCF) at Wright State University.

Perfusion Method

Mouse tissues were obtained from anesthetized (Euthasol 0.1ml/10g, intraperitoneal) animals perfused transcardially with ice-cold vascular rinse solution ( 0.2

M phosphate buffer with 0.8% NaCl, 0.025% KCl, and 0.05% NaHCO3, pH 7.4),

followed by room temperature fixative (4% paraformaldehyde in 0.2 M of phosphate

buffer, pH 7.4). Brain, kidney, liver and skeletal muscle tissues were quickly removed

and post fixed in the same fixative for 2 hours, at 4OC.Then, the tissues were stored

overnight at 4°C in a 15% sucrose solution in 0.1M phosphate buffer for cryoprotection.

Before sectioning tissues were immersed in OCT compound Tissue-Tek® 4583 (Electron

Microscopy Sciences, Hattfield, PA) and mounted on Cryomold® Molds in Adapters

Sakura® Finetek (Sakura Finetec USA INC). Coronal, sagittal or transverse sections 10

µm thickness were cut on a cryostat and mounted on VWR Micro Slides (VWR North

America Cat. No.48312-003). Slides could be stored for months at 4oC.

Antibodies for immunocytochemistry

Immunolabeling of AQP4

Slides with sections mounted were rinsed with PBS-T (0.01 M PBS containing

0.1% Triton-X, pH 7.3) three times for five minutes each, blocked with normal goat serum (10% in PBS) for one hour, and then incubated with primary antibody, overnight, at 4OC. To label AQP4 we used rabbit polyclonal antibody raised against amino acids

244-323 mapping at the C-terminus of AQP4 of human origin (H-80; sc-20812, Santa

Cruz Biotecnology, Inc, Dallas, TX). After several dilution tests, 1:50 dilution was chosen because it gave the clearest results (see table1). The slides were then washed in

0.1M phosphate buffer three times for five minutes each, and incubated with goat anti

rabbit secondary antibody for 45 minutes at room temperature. The secondary antibody

(Alexa Fluor 594-conjugated, affinity purified goat anti-rabbit IgG (Jackson

ImmunoResearch, West Grove, PA) was used at a 1:200 dilution. After incubation with

secondary antibody, the sections were washed again with 0.01M phosphate buffer (three

times for five minutes each), before mounting them in ProLong Gold antifade reagent

with DAPI (Molecular Probes, Life technologies, Ref # P36931). The sections were initially visualized and photographed with an Olympus BX51TF epifluorescence microscope with an attached Spot RT Color camera. Selected sections were visualized and photographed with an Olympus Fluoview, FV1000 confocal microscope.

The specificity of the antibody raised against AQP4 used in this work has been tested in AQP4 KO mouse tissues (Katada et al., 2014), and by Western blotting

(Benfenati et al., 2007). Further, for positive controls for antibody specificity tissues were selected in which AQP4 transcript (Sakai et al., 2014; Mobasheri et al., 2007) and protein

have been demonstrated. Thus, positive controls consisted of immunolabeling of

basolateral membranes of proximal tubule S3 segments in mouse kidney (Van-Hoek et

al., 2000) and mouse skeletal muscle (Frigeri et al., 2004), as shown in Figs. 13B and D.

Negative controls consisted of omitting the primary antibody (Fig.8).

Figure 8. Control of AQP4 immunolabeling in the absence of primary antibody, in mouse kidney cortex. Proximal convoluted tubules and some cortical collecting ducts of mouse kidney were exposed to secondary antibody goat anti-rabbit conjugated to AlexaF-594, without primary antibody. Image obtained from RT-spot camera attached to epifluorescence microscope using a 20X objective. Scale bar represents 20 µm. Background fluorescence (red). Nuclei were stained with DAPI (blue).This image shows that the background fluorescence was minimal and distinctly different from that using primary antibodies (Fig. 13B).

Triple immunolabeling of AQP4, NKCC1 and DAPI

Triple immunolabelling of AQP4, NKCC1 and cell nuclei (DAPI) was done

by sequential immunostaining following a method similar to that described by Wahlby et

al. (2002). First, the slides were rinsed with PBS-T (0.01 M PBS containing 0.1% Triton-

X, pH7.3) three times for five minutes each, blocked with normal goat serum (10% in

PBS; Sigma-Aldrich G 9023) for one hour, and then incubated overnight, at 4oC with

primary anti-AQP4 rabbit polyclonal Ab (H-80; sc-20812). The second day, sections

were washed with 0.1M phosphate buffer three times for five minutes each, and

incubated with goat anti rabbit secondary Ab (Alexa Fluor 594-conjugated goat anti-

rabbit IgG; 1:200, Jackson) for 45 minutes at room temperature. After incubation with

secondary antibodies the sections were washed again with 0.01M phosphate buffer three

times for five minutes each. Then they were incubated with the second primary antibody

i.e. anti-NKCC1 (N-16; sc-21545, affinity purified goat polyclonal antibody raised

against the N-terminus of hNKCC1). Sections were blocked with normal horse serum

(10% in PBS; Life Technologies Cat# 16050-122) for one hour, and then incubated in anti-NKCC1 antibody (1:2000) overnight at 4OC. The third day, the sections were

washed with 0.1M phosphate buffer three times for five minutes each and incubated with

donkey anti rabbit secondary antibody (Alexa Fluor 488-conjugated AffiniPure goat anti-

rabbit IgG, 1:200, Jackson) for 45 minutes at room temperature. After incubation with

secondary antibody, the sections were washed again with 0.01M phosphate buffer three

times for five minutes each before incubation with Prolong Gold antifade reagent with

DAPI (Molecular Probes, Life technologies), overnight at room temperature. All overnight labelling was done by covering the slides with a cardboard box. The specificity of the antibody used in this work against NKCC1 was tested on NKCC1 KO mice choroid plexus sections (Fig.9), where this protein is expressed in the apical membrane

(Piechotta et al., 2002).

Figure 9. NKCC1 antibody specificity test in choroid plexus epithelial cells of lateral ventricle of NKCC1 KO mouse brain. Image shows lack of NKCC1 immunoreactivity. Nuclei stained with DAPI. NKCC1-Ab (1:1000). Secondary antibody donkey anti-goat conjugated to AlexaF-488. Green background fluorescence can be observed. Image from RT-spot camera; 20X objective; scale bar 20 µm. This image should be compared to those in Figs 14 and 15.

Quadruplet immunolabeling of AQP4, NKCC1, GFAP and DAPI

Quadruplet immunolabeling studies were performed by the sequential immunostaining method described above, by treating the sections with antibodies in the following order: AQP4 Ab, NKCC1 Ab, Glial fibrillary acidic protein (GFAP) Ab and

DAPI. Immunolabeling with AQP4 and NKCC1 Abs was exactly as described for triple immunolabeling. The third day Sections were blocked with normal rabbit serum (10% in

PBS) for one hour, and then incubated with anti-GFAP antibody overnight at 4OC. The anti-GFAP (ab4674) affinity purified chicken polyclonal antibody was used at 1:2000 dilutions. The fourth day, the sections were washed in 0.1M phosphate buffer three times for five minutes each and incubated at room temperature, during 45 minutes, with secondary antibodies conjugated to DyLight-649 (rabbit anti-chicken IgG; 1:200, Jackson

ImmunoResearch, West Grove, PA). The sections were washed again with 0.01M phosphate buffer three times for five minutes each before mounting applying Prolong

Gold antifade reagent with DAPI (Molecular Probes, Life technologies) overnight. The specificity of the GFAP Ab used in this study as a glial marker has been validated by

Western blotting (Lonskaya et al., 2013) and numerous immunolabeling studies including astrocytes and radial glial cells (Takano et al., 2006; Achuta et al., 2014).

Immunolabeling of AQP7

The procedure for immunolabeling AQP7 channels was same as described before for AQP4. AQP7 channel immunoreactivity was localized with anti-AQP7 (R-101; sc-

28625) rabbit polyclonal antibody raised against amino acids 169-269 mapping at the C-

terminus of AQP7 of rat origin (Table 1). For AQP7 we used the same blocking serum and secondary antibody used for AQP4. The specificity of the AQP7 antibody used in this work has been tested by preadsorption (Laforenza et al., 2010). Positive controls included detection of immunoreactivity in apical membrane of proximal tubules of mouse kidney (Nejsum et al., 2000) and mouse skeletal muscles (Wakayama et al., 2004), as shown in Figs 16B and C. Negative control by omitting the primary antibody is shown in

Fig. 10. Negative control in a tissue that do not express AQP7 i.e. mouse liver, is shown in Fig. 11. To our knowledge this commercial antibody has not been tested against KO tissues.

Figure 10. Negative control of AQP7 immunolabeling in the absence of primary antibody in mouse kidney cortex. Proximal tubules of mouse kidney treated with secondary antibody goat anti-rabbit conjugated to AlexaF-594. This image from a P30 WT mouse shows slight red background fluorescence, which is very different to that observed when the primary antibody is present as shown in Fig. 16B. Image obtained from RT-spot camera microscope. 40X objective. Scale bar is 10 µm. Nuclei stained with DAPI (blue). S52 P30 36.1.

Figure 11. Negative control of AQP7 antibody in mouse liver cells. A high concentration of AQP primary Ab (1:10) and secondary antibody goat anti-rabbit conjugated to AlexaF-594 (1:200) did not produce IR in liver, which is known to be a tissue where AQ7 is not expressed (Sakai et al, 2014). Image obtained using RT-spot camera. 20X objective; scale bar 20 µm exhibiting slight background fluorescence (red). Nuclei stained with DAPI (blue).

Immunolabeling of AQP9

The procedure for immunolabeling AQP9 channels was basically the same as that

described for AQP4 and AQP7. AQP9 immunoreactivity was localized with anti-AQP9

(AQP91-A) rabbit polyclonal antibody raised against 18 amino acids mapping at the C-

terminus of rat AQP9 (Table 1).The same blocking serum and secondary antibody used

for AQP4 and 7 was used for AQP9. The specificity of the antibody has been tested

against AQP7 KO tissues (Rojek et al., 2007). Positive controls of the antibody consisted

of quantitative real-time PCR analysis (Sakai et al., 2014) and immunoreactivity in liver

cells (Elkjaer et al., 2000) (see figure of results obtained). Negative Control consisted of

staining in kidney cells with known no AQP9 expression (Ishibashi et al., 1998) (Fig. 12).

Triple immunolabeling studies were also performed by sequential immunostaining

method with AQP9 antibody, GFAP antibody and nuclei stained with DAPI in same procedure described before for AQP4 triple immunolabeling.

Figure 12. Negative control of AQP9 antibody in mouse kidney. AQP 9mRNA is not expressed in mouse kidney (Sakai et al, 2014). Thus this tissue has been used a control for immunocytochemistry (Badaut et al, 2001, Nielsen et al, 2001). In this section of proximal tubule cells of mouse kidney were immunostained using AQP9 antibody (1:500). Secondary antibody was goat anti-rabbit conjugated to AlexaF-594. Image obtained from RT-Spot camera attached to epifluorescence microscope. Objective 20X. Scale 20 µm. Note background fluorescence (red). Nuclei stained with DAPI (blue).

Epifluorescence microscopy and confocal imaging

Images were initially obtained on an Olympus Epi-fluorescence Spot scope with an RT color camera. The epifluorescence microscope (Olympus BX51) attached to the

Spot camera was equipped with the following objectives, all of which were used at

various stages of this work: 4X objective (NA 0.16), 10X objective (NA 0.26), 20X objective (NA 0.75) and 60X oil immersion (NA 1.35). Confocal images were obtained from a Fluoview 1000 Olympus microscope, using a 60X (NA 1.35) oil immersion objective at 0.25 µm Z-steps. Molecular labels were excited using the following laser lines (in nm): 405 (DAPI), 488 (NKCC1), 561 for AQP4,7 and 9; and 635 for GFAP.

Confocal optical sections were collected and analyzed using Fluoview software. Stacks from confocal images were generated using ImageJ software (imagej.nih.gov/ij), using

TIFF images exported from Fluoview Software. Images were added to the “stack” option in the main toolbar of Image J software. The only image stacks shown in this work are in

Figs. 15A and 16 A.

Table1. Primary Antibodies used in the present study

PRIMARY HOST SOURCE TESTED DILUTION ANTIBODY TYPE AGAINST KO MOUSE

Anti-AQP4 Rabbit Santa Cruz Yes 1:50 (H-80) Polyclonal Biotechnology, sc-20812 Inc.

Anti-AQP7 Rabbit Santa Cruz No 1:50 (R-101) Polyclonal Biotechnology, sc-28625 Inc.

Anti-AQP9 Rabbit Alpha Yes 1:500 & 1:200 (AQP91-A) Polyclonal Diagnostic International

AntiNKCC1 Goat Santa Cruz Yes 1:2000 (N-16) Polyclonal Biotechnology, sc-21545 Inc.

Anti-GFAP Chicken Abcam Inc. No 1:2000 (ab4674) Polyclonal

RESULTS

Localization of Aquaporin 4 in mouse choroid plexus epithelium and ependymal cells

Immunolabeling methods with validated antibodies and confocal immunofluorescence were used to determine the cellular localization of AQP4 within the choroid plexus epithelium and ependymal cells. AQP4 was not found in the choroid plexus (CP) but in the epedymal region (EP) lining the lateral ventricles. The AQP4-IR in

EP was found in the basal aspect of the ependymal cells, and in finger-like structures in between the EP cells. These AQP4-immunoreactive structures having calyx-like shape, surounding the basolateral aspects of EPCs are shown at higher maginification in Fig.

13D (arrowheads). AQP4-IR was also prominent in structures surrounding capillaries as shown in Fig. 13A (CAP). As positive controls for antibody specificity we used kidney proximal tubules (Fig. 13B) and skeletal muscle sarcolemma (Fig. 13C). AQP4 expression in the basolateral membrane of proximal kidney tubules is well established both in both mouse (Van-Hoek et al., 2000) and rat (Huang et al., 2001). Similarly, AQP4 expression in skeletal muscle is well documented (Yang et al., 2000; Frigeri et al., 1995,

2004).

Figure 13. Aquaporin 4 water channel protein (AQP 4) immunolabeling in mouse (P-30) lateral brain ventricle structures, kidney proximal tubules and skeletal muscle fibers. A) Micrograph of a confocal optical section (0.25 µm) of the lateral ventricle (LV) showing AQP4 immunoreactivity (IR) in ependyma (EP) and surrounding an arteriolar capillary (CAP). Note that the choroid plexus (CP) does not show AQP4-IR. Strong AQP4-IR is found in the basal aspect of ependymal cells, and in between these cells. The framed area in A is shown at higher magnification in D, where arrowheads indicate the AQP4 immunoreactive structures, which appeared encroaching the ependymal cells. B) Non-confocal immunolabeling of kidney proximal tubules showing basolateral AQP4 location (arrows). C) Non-confocal immunofluorescence of hind limb skeletal muscle fibers where AQP4-IR is evident in the sarcolemma. AQP4-IR is shown in red (Alexa F 594 secondary Ab, primary anti-AQP 4 polyclonal Ab sc-20812). Nuclei stained with DAPI (blue). A 60X oil objective was used to obtain images in A and D. B and C immunofluorescence was visualized with a 20X objective. Scale bars in A and D, 10 µm; and in B and C, 20 µm.

Colocalization of AQP4 and NKCC1 in mouse ependyma and capillaries

The AQP4-IR shown in the EP (Fig. 13 A and D; Fig. 14D) could be located in

the basolateral membrane of ependymal cells, or in structures surrounding these cells. To

investigate this issue, we studied the cellular and subcellular colocalization of AQP4 with

NKCC1, the Na+,K+, 2Cl- cotransporter known to be expressed in astrocytes (Plotkin et

al, 1997; Yan et al, 2001, 2001) and in the apical membrane of CPECs (Plotkin et al.,

1997; Piechotta et al 2002). As expected, NKCC1 was found to be expressed in the

apical membrane of CPECs (Fig. 14 A). NKCC1 was expressed not only in CPECs but in

the apical membrane of ependymocytes (EP in Fig. 14 A, green and C green). NKCC1

was also expressed in the basal aspect of the ependymal cells (subependymal zone) where

it colocalized with AQP4 (Fig. 14 A, yellow, and Fig. 14D yellow). We also found very

clear colocalization of NKCC1 and AQP4 in structures surrounding capillaries (asterisk

in Fig. 14A, yellow). These structures most likely represent astrocytes endfeet, since they

were labelled with GFAP (not shown). Thus, the data confim NKCC1 expression in the apical membrane of CPECs and shows for the first time that NKCC1 is also expressed in the apical membrane of lateral ventricle ependymal cells. As for the basal location of

NKCC1 in between ependymal cells and in their basal aspect in the subependymal zone, it was found that this cotransporter colocalized with AQP4. The question arises as to whether NKCC1 and AQP4 in the basolateral aspect of ependymal cells were located in the endings of radial glial cells known to be present in between ependymal cells and in their basal pole, in thesubependymal zone (Götz, 2013), or in the basolateral membrane

of the EP cells. To investigate this, we used quadruplet immunolabelling of NKCC1,

GFAP, AQP4 and DAPI. The results are shown in Fig. 15.

See figure legend on next page

Figure 14. Confocal images of triple immunolabeling of AQP4 (red), NKCC1(green) and DAPI (blue) in choroid plexus epithelial cells (CPECs) and ependyma (EP) of lateral ventricle (LV) of mouse (P-30) brain. A) NKCC1 immunolabeling (green) in the apical membrane of both CPECs and ependymal cells (EP). NKCC1 colocalized with AQP4 (yellow merge) in structures in between ependymal cells and in their basal aspect in the subependymal zone. Similarly, NKCC1 and AQP4 were colocalized in a subependymal arteriolar capillary. The framed area in A is shown at higher magnification in B, C, and D. B) AQP4 (red) immunoreactivity in between ependymal cells and in their basal poles in the subependymal zone. Arrowhead shows one intercellular AQP4 immunoreactive structure. Most of these structures coexpressed NKCC1, as can be observed in A and D (yellow). C) High magnification NKCC1-IR in the apical membrane of CPECS and EP cells. D) Merged image of B and C to show colocalization (yellow) of NKCC1 and AQP4. Note that apical NKCC1 in CPECs and EP cells does not colocalize with AQP4. NKCC1 and AQP4 colocalized in structures in between ependymal cells and in the subependymal pole of these cells. Images were obtained with a 60X Oil Objective at 0.25 um Z-step; scale bars represent 10 µm. AQP4 secondary Ab (red, Alexa F 594), Anti-AQP 4 primary Ab (sc-20812), NKCC1-IR secondary Ab (green, Alexa F 488). Anti-NKCC1 primary Ab (sc-21545). Nuclei stained with DAPI (blue).

Colozalization of AQP4, NKCC1 and GFAP in ependyma and subependymal layers.

To investigate if the structures expressing AQP4 and NKCC1 were glial processes, we used quadruplet immunolabeling of sections of mouse brain stained with antibodies against AQP4, NKCC1, GFAP and DAPI (Fig. 15). The results confirmed strong NKCC1-IR (in blue) in the apical membrane of CPECs, and in the apical domain of the ependymal cells (EP) throughout the lining of the lateral ventricles (Fig. 15A, B).

NKCC1 in CPECs and EP cells was not colocalized with GFAP or AQP4. In contrast,

NKCC1 expressed in structures surrounding the basolateral aspects of ependymal cells colocalized with GFAP, which in the merge images appears in cyan (Fig. 15 A and B, arrowheads). The details of the framed area in Fig. 15A are shown in Figs. 15B-D at higher magnification. Fig. 15B, shows the co-localization of NKCC1-IR (blue) and

GFAP-IR (green), which appears in cyan (arrows), suggesting that the subependymal structures end processes of radial glia. AQP4 co-localized with GFAP and it is also found in the “end-feet” of radial glial structures (Fig. 15C, yellow). Fig. 15D shows the

expression of GFAP-IR (green) and DAPI (white) in the subependymal zone and in between EPCs.

Figure 15. Confocal images of quadruplet immunolabeling of NKCC1, AQP4, GFAP and DAPI in choroid plexus epithelial cells (CPECs) and ependyma (EP) of lateral ventricles (LV) of mouse (P-30) brain. A) NKCC1 (blue) is expressed in the apical membrane of ependymal cells (EP) and in the apical membrane of CPECs (CP). NKCC1 in the basal aspect of the EP colocalized with GFAP (green); the merge image appears in cyan. The square framed are in A, is shown at higher magnifications in B, C and D. B) colocalization of NKCC1 (blue) and GFAP (green) in cyan (arrow heads) in between ependymal cells and subependymal layer structures. Note apical NKCC1 in ependymal cell. C) Colocalization (yellow) of AQP4 (red) in GFAP (green) immunoreactive structures in the subependymal zone. D) Expression of glial marker GFAP in subependymal structures surrounding the cell bodies of ependymal cells. DAPI in all panels is in white. Images are stacks of 4 confocal optical sections (slices Z27-Z30) at 0.25 µm steps in the z axis, taken with a 60X oil objective. Scale bars 10µm. All Abs as in previous figures, except for GFAP secondary Ab (green, Dylight 649) and primary anti-GFAP pAb ab4674.

Localization of Aquglyceroporin 7 (AQP7) in mouse choroid plexus and ependymal cells Fig. 16A shows a mouse brain at the level of the lateral ventricle. The section was

labelled with AQP7 antibody. We have found distinct immunofluorescence restricted to

the apical membrane of CPECs. AQP7 expression was also found in the apical membrane

of ependymal cells. As positive control we used mice tissues where AQP7 expression is

well established i.e. apical membrane of kidney proximal tubules and skeletal muscle. A clear pattern of AQP7 IR (arrows) is observed in Fig. 16B, the apical membrane of

proximal convoluted tubules of mouse kidney. Fig 16C, shows strong AQP4-IR (arrows)

in sarcolemma of skeletal muscle fibers.

See figure legend on next page

Figure 16. Aquaporin 7 water channel protein (AQP 7) immunolabeling in mouse (P-30) lateral brain ventricle structures, kidney proximal tubules and skeletal muscle fibers. A) Micrograph of a confocal optical stack prepared from slices Z31-Z38, each slice (0.25 µm) of the lateral ventricle (LV) showing AQP7 immunoreactivity (IR) in the apical membranes of Choroid plexus (CP) and ependyma (EP). B) Non-confocal immunolabeling of kidney proximal tubules showing apical AQP7 localization (arrows). C) Non-confocal immunofluorescence of limbs skeletal muscle fibers where AQP7-IR is evident in the sarcolemma. AQP7-IR is shown in red (Alexa F 594 secondary Ab, primary anti-AQP 7 polyclonal Ab sc-28625). Nuclei stained with DAPI (blue). A confocal 60X oil objective was used to obtain images in A. B and C immunofluorescence was visualized with a 20X objective non confocal microscope. Scale bars in A, 10 µm; and in B and C, 20 µm.

Localization of Aquglyceroporin 9 (AQP9) in mouse choroid plexus and ependymal

cells

Fig 17 A, B and C shows mouse brain at the level of thalamus and the auditory

neocortex. The section was labelled with AQP9 antibody. AQP9 immunofluorescence

was found absent in CP, but was found restricted in the sub ependymal zone of some

areas of the lateral ventricle, specifically in the areas surrounding the thalamus and the

neocortex (Fig. 17 A), where it colocalized with GFAP (Fig. 17 C). AQP9-IR was

colocalized with GFAP (glial marker) suggesting its glial location. Fig. 17 B shows the

GFAP-IR in subependyma. As positive control we used mouse tissue where AQP9

expression is well known and undisputed, like mouse liver cells and glial cells. Fig. 17 D shows strong AQP9-IR (arrows) in the apical and few domains of basolateral membrane of bile duct of liver. Fig. 17 E, F shows strong AQP9-IR and GFAP-IR between fimbria and lining of lateral ventricle The Co-localization of AQP9 and GFAP-IR is in yellow between fimbria and lining of ventricle (Fig. 18G).

See figure legend on next page

Figure 17. Aquaporin 9 water channel protein (AQP 9) immunolabeling in mouse (P-30) brain thalamus, the auditory neocortex and liver bile duct. A) Non-confocal immunolabeling of AQP9 showing that there is no AQP9 immunoreactivity (IR) of AQP9 in CPECs and in Ependyma (EP) but strong IR (arrows) is observed in the subependymal layer surrounding the thalamus and the auditory neocortex. B) Shows GFAP-IR in subependyma. C) Shows Co-localization of AQP9 and GFAP IR in yellow. D) Shows strong AQP9-IR (arrows) apical and basolateral membrane of bile duct of liver. E) Strong AQP9-IR between fimbria and lining of ventricle. F) Strong GFAP-IR between fimbria and lining of ventricle. G) Co-localization of AQP9 and GFAP IR in yellow between fimbria and lining of ventricle. AQP9-IR is shown in red (Alexa F 594 secondary Ab, primary anti-AQP 9 polyclonal Ab AQP91-A). Nuclei stained with DAPI (blue). Immunofluorescence was visualized with a 20X objective non confocal microscope in images in A, B and C. Non confocal 60X oil objective was used to obtain images in D, E, F and G. Scale bars in A, B, C, 20 µm; and in D, E, F, G, 10 µm.

DISCUSSION

The CP secretes most of the CSF though mechanisms that are poorly understood

(Damkier et al., 2013). The process of fluid secretion in CP involves transcellular water

transport through CPECs, but the molecular nature of these water pathways is an unresolved matter. As already mentioned in the introduction of the present work, water

flux via the lipid bilayers cannot explain the high rates of fluid secretion typical of

CPECs. Two major cation-coupled cotransporters that “cotransport” water and ions are expressed in these cells, NKCC1 and KCC3a. However, under basal conditions these cotransporters work in the opposite direction to that expected for fluid secretion (see

Fig.1). Another cation-coupled cotransporter, KCC4, which has the expected orientation for solute and fluid secretion (Fig. 1), is activated by osmotic cell swelling and thus, it is not clear if it is involved in water cotransport (reviewed in Alvarez-Leefmans, 2012).

Thus, the most likely candidates for fluid secretion in CPECs are AQPs.

Net water flux occurs from the basolateral membrane through the apical membrane of CPECs (Alvarez-Leefmans, 2012; Damkier et al., 2013). Thus, elucidation

of the expression and localization of AQPs in CPECs constitutes a crucial step for

understanding the molecular machinery involved in CSF secretion. However, with the

exception of AQP1 in the apical membrane, the molecular pathways for transepithelial water transport are either unknown or their expression is controversial; there are no known AQPs in the basolateral membrane of CPECs to account for the influx of water, and whether AQPs other than AQP1 are expressed in the apical membrane remains to be

elucidated. The expression of AQP1 in the basolateral membrane is a subject of debate.

Previous work from our lab confirmed the strict apical localization of AQP1; its expression in the basolateral membrane of mouse CPECs is undetectable (Crum and

Alvarez-Leefmans, 2012). Thus, AQPs other than AQP1 are potential molecular substrates for the basolateral water influx. Further, it is not known if besides AQP1 there are other AQPs in the apical membrane that could participate in CSF secretion.

In the present study, we began testing the hypothesis proposing that there must be

AQPs other than AQP1 localized to the apical membrane and to the basolateral

membrane. AQPs in the basolateral membrane could explain basolateral net water influx.

Previously, our group detected the transcripts for several AQPs in whole choroid plexus tissue using RT-PCR (Cha et al., 2014). Those results are only a preliminary step because

of possible contamination with ependymal and or glial cells. Nevertheless they provided the clues for investigating the expression of various water channel proteins in CPECs and

EPCs. The AQPs identified at the transcript level were AQP0, AQP1, AQP2, AQP4, and aquaglyceroporins AQP3, AQP7 and AQP9, and “unorthodox” AQP 11. The present

work was restricted to study the expression pattern of three of these AQPs: AQP4, AQP7

and AQP9.

Expression of AQP4

The present work constitutes a detailed study of the cellular and subcellular

distribution of AQP4 in CPECs and EPCs of LVs of mouse brain, using confocal

immunofluorescence with antibodies validated against KO tissues (Katada et al., 2014)

and positive controls with tissues in which AQP4 expression is undisputed. In mice of

different ages (P-12, 19, 20, 29 and 30) AQP4 was not expressed in CPECs (Fig. 13A),

but rather in radial glial endings surrounding the basolateral aspect of EPCs (Fig. 15C),

and in astrocytic end-feet surrounding blood vessels (Fig. 13A and 14A). The

expression of AQP4 in CPECs has been a controversial topic; different research groups report conflicting results using in situ hybridization (Hasegawa et al., 1993; Jung et al.,

1994, Venero et al, 1999) or immunocytochemistry (Nielsen et al., 1997; Speake et al.,

2003; Mobasheri et al., 2007; Nazari et al., 2015). The present results are consistent with

in situ hybridization studies suggesting the absence of mRNA encoding AQP4

(Hasegawa et al., 1993; Jung et al., 1994) but do not support similar observations by

Venero’s group (Venero et al., 1999). As for AQP4 protein expression, our results are in

agreement with those of Nielsen’s group (Nielsen et al., 1997) but in sharp disagreement

with studies proposing AQP4 protein expression in CPECs (Speake et al., 2003;

Mobasheri et al., 2007; Nazari et al., 2015). The antibodies used in the present study

were validated against AQP4 -/- mouse tissues (Katada et al., 2014), and as positive

controls we used mouse tissues where AQP4 expression is undisputed, such as the

basolateral membranes of proximal tubules (Van Hoek et al., 2000) and in mouse skeletal

muscle (Frigeri et al., 1995; Yang et al., 2000; Crosbie et al., 2002). Previous work for

our group (Cha et al, 2014) detected AQP4 mRNA transcripts by RT-PCR analysis of

whole CP tissue. This result is likely to be explained by contamination with ependymal or

glial cells, during extraction of the CP from the IV ventricle of mouse brain.

We have found the expression of AQP4 in endings of radial glial cells (glia

limitans?) and end feet of perivascular astrocytes (Fig. 15C), which is in agreement with

previous studies (Frigeri et al., 1995; Rash et al., 1998). For review see (Nagelhus and

Ottersen, 2013, Papadopoulos and Verkman, 2013). This specific distribution of AQP4

suggests that it plays an important role in water (and ion) transport at the brain–CSF

interface, i.e. the ependyma, perhaps in K+ clearance from CSF. This intriguing possibility motivated us to further examine the coexpression of AQP4 and NKCC1 in ependymal and subependymal regions, given that NKCC1 has properties that make it suitable as an extracellular K+ uptake mechanism (Alvarez-Leefmans, 2012).

Co-expression of AQP4 and NKCC1

Given the distribution of AQP4-IR in glial processes ending between and around

ependymal cells we searched for the possible coexpression with NKCC1. We

hypothesized that NKCC1 might be colocalized with AQP4 not only in the astrocytic

perivascular endfeet but also in the ventricular endings of radial glia. The results show

that AQP4 colocalized with NKCC1 in both the radial glial endings surrounding the basal

aspect of EPCs (Fig 15A, B and C), and in astrocytic perivascular endfeet (Figs 14A and

15A). Interestingly, we also found the expression of NKCC1 in the apical membrane of

EPCs (Fig 14A and 15A). The antibodies used in this study were validated against

NKCC1 -/- mouse tissues, and as positive controls we used choroid plexus epithelium

itself. In epithelial cells NKCC1 is expressed in the basolateral membrane except for

CPECs where it adopts an apical location. To the list of NKCC1 expression exceptions in

epithelial cells we now have to add the brush border, apical membrane of ependymal

cells where it may play a role in cell volume maintenance and K+ buffering of CSF and

brain interstitial fluid.

Radial glial cells, astrocytes and CPECs are believed to play a significant role in

K+ uptake in CNS (Abbott et al, 2006; Nagelhus and Ottersen, 2013) and CSF

respectively (Wu et al., 1998; Alvarez-Leefmans, 2012). We found colocalization of

NKCC1 and glial fibrillary acidic protein in radial glial endings surrounding EPCs and in

astrocytes of cortex, hippocampus and corpus callosum of mouse brain, consistent with

previous studies showing coexpression of NKCC1 with glial cells (Yan et al 2001a,b). It

has long been proposed that astrocytes and other glial cells have a major role in buffering

extracellular K+ by transferring these ions released by active neurons in the CNS to blood.

Thus, glial cells uptake excess extracellular K+, which can be transported to other

locations where extracellular K+ concentrations are lower or these ions are released at the

endfeet of perivascular astrocytes forming the BBB, where they are transported to blood

vessels (Paulson and Newman, 1987).

What is the functional significance of these findings? Two recent papers addressing the functional aspects of the proteins thought to be involved in K+ clearance,

coauthored by Nana MacAulay (Macaulay and Zeuthen, 2012; Larsen et al, 2014) serve

as example of the current controversies on the functional role of the triad formed by

AQP4, inward rectifier K+ channels (Kir) and NKCC1 on extracellular K+ buffering by

astrocytes. They postulated two different views that are contradictory: i) K+ clearance from the extracellular space is likely to be achieved by the coordinated action of AQPs

and NKCC1 (Macaulay and Zeuthen, 2012). ii) Later, in 2014, Larsen and coworkers

claimed that NKCC1 does not contribute to K+ clearance and buffering, although the

results they presented were not convincing (Larsen et al, 2014). The controversy about the role of NKCC1 in conjunction with AQP4 in K+ uptake in the CNS is further illustrated by two letters to the Editor published in the Journal of General Physiology in

2013, one supporting the role of NKCC1 (Hertz et al, 2013) and the other challenging it

(Jin et al, 2013). The present findings showing AQP4 and NKCC1 co-expression strongly suggest a role of NKCC1 in K+ uptake from the cerebrospinal fluid, thus buffering and regulating K+ in CSF and brain interstitial fluid.

The schematic diagram of Fig. 18 illustrates the distribution of AQP4 and

NKCC1 in the endings of radial astrocytes and in EPCs. AQP4 and NKCC1 are present in the endings of radial astrocytes (green) forming calyx-like structures around the basolateral aspect of ependymal cells (yellow). Also, illustrated in Fig. 18 is the expression of NKCC1 in the apical membrane of ependymocytes.

Figure 18. Diagram illustrating the distribution of AQP4 and NKCC1 in the endings of radial glia and in the ependyma. The drawing summarizes the results obtained by immunolabeling experiments aimed at determining the sub-cellular localization of AQP4 and NKCC1 colocalized with GFAP. AQP4 and NKCC1are present in the endings of glial cells, probably radial glia or tanycytes (green), which are encroaching the ependymal cells (yellow). The drawing also illustrates the expression of NKCC1 in the apical membrane of EPCs. The dyad AQP4-NKCC1 may play a significant role K+ regulation of CSF.

Expression of AQP7 The studies aimed at determining the subcellular distribution of AQP7 in CPECs and EPCs showed that AQP 7 expression is restricted to the apical membranes of CPECs

and EPCs. (Fig. 16A). The antibodies used in this work have been validated in previous

studies (Laforenza et al., 2010, 2013; Goubau et al., 2013; Hermo et al., 2008; Aure et al.,

2014). As positive controls we used mouse tissues where AQP7 expression is well established like mouse skeletal myofibers (Wakayama et al 2014) and apical (brush

border) membrane of the proximal tubules of mouse kidney (Nejsum et al, 2000).

Our results are in agreement with a previous study claiming AQP7 expression in

CPECs but its specific subcellular localization i.e. apical or basal membrane on CPECs

was not discerned Shin et al, 2006). The present study shows, for the first time, the

subcellular localization of AQP7 in CPECs. AQP7 is largely expressed in the apical membrane of CPECs and in varying degree of in ependyma of LV of mouse brain. AQP7 was undetected in basolateral membrane of CPECs. This suggests that AQP7 could work together with AQP1 in CSF secretion in CP.

Expression of AQP9

We present a detailed immunolabeling and confocal immunofluorescence study of the distribution of AQP9 in CPECs and EPCs of LV mouse brain in seven different mice of ages P-12, 19, 20, 29 and 30. AQP9 is expressed neither in apical nor in basolateral membrane of CPECs (Fig. 17A). Thus, the water pathways for water influx across the

CPECs basolateral membrane remain to be elucidated. However, we have found the expression of AQP9 in radial glial cells and in the subependymal zone (Fig. 18 B, C).

The antibodies used in this study were validated against AQP9 -/- mouse tissues

(Mylonakou et al., 2009), and as positive controls we used mouse liver tissues where

AQP9 expression was established previously (Nicchia et al., 2002; Badaut et al., 2004).

Our findings on AQP9 expression in radial glial (Fig. 18E, F and G) and tanycytes are in agreement with previous studies (Elkjaer et al., 2000, Badaut et al., 2001).

This specific distribution of AQP9 suggests that it may play a role in water transport at brain–CSF and brain– blood interfaces.

CONCLUSIONS

1) Validated antibodies against AQP4 revealed that this protein is not expressed in

CPECs.

2) Ependymal AQP4 is not expressed in ependymocytes but in the endings of radial glia

surrounding the basal aspect of the ependymal cells. In this location AQP4 is co-

expressed with NKCC1.

3) The data confirm NKCC1 expression in the apical membrane of CPECs and shows for

the first time that NKCC1 is also expressed in the apical membrane of lateral ventricle

ependymal cells.

4) The expression of NKCC1 in the apical membrane of EPCs and its colocalization with

AQP4 at the ventricular endings of radial glia suggest NKCC1 could play a role in CSF

and brain interstitial fluid K+ homeostasis.

5) AQP7 is expressed in the apical membrane of CPECs and EPCs suggesting that together with AQP1 may be involve in CSF secretion.

6) AQP9 was found in the subependymal region coexpressed with the glial marker

GFAP. AQP9 is expressed neither in CPECs nor in EPCs.

7) The AQP expected to be expressed in the basolateral membrane of CPECs remains to be elucidated.

REFERENCES

Abbott, N. J. (2004). Evidence for bulk flow of brain interstitial fluid:

Significance for physiology and pathology. Neurochemistry International, 45(4), 545-

552. doi:10.1016/j.neuint.2003.11.006 [doi]

Achuta, V. S., Rezov, V., Uutela, M., Louhivuori, V., Louhivuori, L., & Castren,

M. L. (2014). Tissue plasminogen activator contributes to alterations of neuronal migration and activity-dependent responses in fragile X mice. The Journal of

Neuroscience : The Official Journal of the Society for Neuroscience, 34(5), 1916-1923. doi:10.1523/JNEUROSCI.3753-13.2014 [doi]

Albertini, R., & Bianchi, R. (2010). Aquaporins and glia. Current

Neuropharmacology, 8(2), 84-91. doi:10.2174/157015910791233178 [doi]

Alvarez-Leefmans, F. J. (2012). Intracellular Chloride Regulation. Cell

Physiology Sourcebook: Essentials of Membrane Biophysics, 221-259.

Anderson, J. M., & Van Itallie, C. M. (1995). Tight junctions and the molecular basis for regulation of paracellular permeability. The American Journal of Physiology,

269(4 Pt 1), G467-75.

Angelow, S and Galla, H.J. (2005) Junctional Proteins of the Blood-CSF Barrier.

Ch. 3 in: The Blood Cerebrospinal Fluid Barrier, Zheng, W. and Chobdobsky, A. 53-83.

Arany, S., Benoit, D. S., Dewhurst, S., & Ovitt, C. E. (2013). Nanoparticle- mediated gene silencing confers radioprotection to salivary glands in vivo. Molecular

Therapy : The Journal of the American Society of Gene Therapy, 21(6), 1182-1194. doi:10.1038/mt.2013.42 [doi]

Aure, M. H., Ruus, A. K., & Galtung, H. K. (2014). Aquaporins in the adult mouse submandibular and sublingual salivary glands. Journal of Molecular Histology,

45(1), 69-80. doi:10.1007/s10735-013-9526-3 [doi]

Badaut, J., Hirt, L., Granziera, C., Bogousslavsky, J., Magistretti, P. J., & Regli,

L. (2001). Astrocyte-specific expression of aquaporin-9 in mouse brain is increased after transient focal cerebral ischemia. Journal of Cerebral Blood Flow and Metabolism :

Official Journal of the International Society of Cerebral Blood Flow and Metabolism,

21(5), 477-482. doi:10.1097/00004647-200105000-00001 [doi]

Badaut, J., Petit, J. M., Brunet, J. F., Magistretti, P. J., Charriaut-Marlangue, C., &

Regli, L. (2004). Distribution of aquaporin 9 in the adult rat brain: Preferential expression in catecholaminergic neurons and in glial cells. Neuroscience, 128(1), 27-38. doi:10.1016/j.neuroscience.2004.05.042 [doi]

Barry, D. S., Pakan, J. M., & McDermott, K. W. (2014). Radial glial cells: Key organisers in CNS development. The International Journal of Biochemistry & Cell

Biology, 46, 76-79. doi:10.1016/j.biocel.2013.11.013 [doi]

Beitz, E. (2009) Aquaporins Handbook of Experimental Pharmacology 190, http://www.springer.com/us/book/9783540798842#aboutBook

Benfenati, V., Nicchia, G. P., Svelto, M., Rapisarda, C., Frigeri, A., & Ferroni, S.

(2007). Functional down-regulation of volume-regulated anion channels in AQP4 knockdown cultured rat cortical astrocytes. Journal of Neurochemistry, 100(1), 87-104. doi:JNC4164 [pii]

Brightman MW. (1965) The distribution within the brain of ferritin injected into cerebrospinal fluid compartments. I. Ependymal distribution. J Cell Biol. Jul; 26(1):99-

123. PubMed PMID: 5859025; PubMed Central PMCID: PMC2106700.)

Brown, P. D., Davies, S. L., Speake, T., & Millar, I. D. (2004). Molecular mechanisms of cerebrospinal fluid production. Neuroscience, 129(4), 957-970. doi:S0306-4522(04)00564-0 [pii]

Calamita G, Gena P, Ferri D, Rosito A, Rojek A, Nielsen S, Marinelli RA,

Fruhbeck G & Svelto M (2012) Biophysical assessment of aquaporin-9 as principal facilitative pathway in mouse liver import of glucogenetic glycerol. Biology of the Cell

104 342–351. (doi:10.1111/boc.201100061)

Cavazzin, C., Ferrari, D., Facchetti, F., Russignan, A., Vescovi, A. L., La Porta,

C. A., et al. (2006). Unique expression and localization of aquaporin-4 and aquaporin-9 in murine and human neural stem cells and in their glial progeny. Glia, 53(2), 167-181. doi:10.1002/glia.20256 [doi]

Cha, D., - Crum, J., & - Alvarez-Leefmans, F. - Molecular pathways for water fluxes in mouse choroid plexus epithelial cells (1182.3) The FASEB Journal vol. 28 no. 1

Supplement 1182.3, 2014

Chikly, B., & Quaghebeur, J. (2013). Reassessing cerebrospinal fluid (CSF) hydrodynamics: A literature review presenting a novel hypothesis for CSF physiology.

Journal of Bodywork and Movement Therapies, 17(3), 344-354. doi:10.1016/j.jbmt.2013.02.002 [doi]

Cifuentes M, Fernández-LLebrez P, Pérez J, Pérez-Fígares JM, Rodríguez EM.

Distribution of intraventricularly injected horseradish peroxidase in cerebrospinal fluid compartments of the rat spinal cord. Cell Tissue Res. 1992 Dec;270(3):485-94. PubMed

PMID: 1486601.

Crosbie, R. H., Dovico, S. A., Flanagan, J. D., Chamberlain, J. S., Ownby, C. L.,

& Campbell, K. P. (2002). Characterization of aquaporin-4 in muscle and muscular dystrophy. FASEB Journal : Official Publication of the Federation of American Societies for Experimental Biology, 16(9), 943-949. doi:10.1096/fj.01-0327com [doi]

Crum, J. M., Alvarez, F. J., & Alvarez-Leefmans, F. J. (2012). The Apical

NKCC1 Cotransporter Debate. FASEB Journal, 26:881-884.

Damkier, H. H., Brown, P. D., & Praetorius, J. (2010). Epithelial pathways in choroid plexus electrolyte transport. Physiology (Bethesda, Md.), 25(4), 239-249. doi:10.1152/physiol.00011.2010 [doi]

Damkier, H. H., Brown, P. D., & Praetorius, J. (2013). Cerebrospinal fluid secretion by the choroid plexus. Physiological Reviews, 93(4), 1847-1892. doi:10.1152/physrev.00004.2013 [doi]

Dandy, W. E. (1919). Experimental hydrocephalus. Annals of Surgery, 70(2),

129-142.

Elkjaer, M., Vajda, Z., Nejsum, L. N., Kwon, T., Jensen, U. B., Amiry-

Moghaddam, M., et al. (2000). Immunolocalization of AQP9 in liver, epididymis, testis, spleen, and brain. Biochemical and Biophysical Research Communications, 276(3),

1118-1128. doi:10.1006/bbrc.2000.3505 [doi]

Finkelstein, A. (1976). Water and nonelectrolyte permeability of lipid bilayer membranes. The Journal of General Physiology, 68(2), 127-135.

Fischbarg, J. (2010). Fluid transport across leaky epithelia: Central role of the tight junction and supporting role of aquaporins. Physiological Reviews, 90(4), 1271-

1290. doi:10.1152/physrev.00025.2009 [doi]

Friedman, M. H. (2008). Epithelial Transport. Chapter 10 in: Principles and

Models of Biological Transport; pages 391-446, Springer, NY

Frigeri, A., Gropper, M. A., Turck, C. W., & Verkman, A. S. (1995).

Immunolocalization of the mercurial-insensitive water channel and glycerol intrinsic protein in epithelial cell plasma membranes. Proceedings of the National Academy of

Sciences of the United States of America, 92(10), 4328-4331.

Frigeri, A., Nicchia, G. P., Balena, R., Nico, B., & Svelto, M. (2004). Aquaporins in skeletal muscle: Reassessment of the functional role of aquaporin-4. FASEB Journal :

Official Publication of the Federation of American Societies for Experimental Biology,

18(7), 905-907. doi:10.1096/fj.03-0987fje [doi]

Fruhbeck G (2005) Obesity: aquaporin enters the picture. Nature 438-436–437.

(doi:10.1038/438436b)

Fruhbeck G, Catalan V, Gomez-Ambrosi J & Rodriguez A 2006 Aquaporin-7 and glycerol permeability as novel obesity drug-target pathways. Trends in Pharmacological

Sciences 27 345–347. (doi:10.1016/j.tips.2006.05.002)

Götz, M (2013). Radial Glial Cells. Ch 5 in : Neuroglia, third edition,

Kettenmann, H., ransom, B.R. Oxford.

Gunzel, D., & Yu, A. S. (2013). Claudins and the modulation of tight junction permeability. Physiological Reviews, 93(2), 525-569. doi:10.1152/physrev.00019.2012

[doi]

Goubau, C., Jaeken, J., Levtchenko, E. N., Thys, C., Di Michele, M., Martens, G.

A., et al. (2013). Homozygosity for aquaporin 7 G264V in three unrelated children with hyperglyceroluria and a mild platelet secretion defect. Genetics in Medicine : Official

Journal of the American College of Medical Genetics, 15(1), 55-63. doi:10.1038/gim.2012.90 [doi]

Hakvoort, A., Haselbach, M., Wegener, J., Hoheisel, D., & Galla, H. J. (1998).

The polarity of choroid plexus epithelial cells in vitro is improved in serum-free medium.

Journal of Neurochemistry, 71(3), 1141-1150.

Hamann, S., Herrera-Perez, J. J., Bundgaard, M., Alvarez-Leefmans, F. J., &

Zeuthen, T. (2005). Water permeability of Na+-K+-2Cl- cotransporters in mammalian

epithelial cells. The Journal of Physiology, 568(Pt 1), 123-135. doi:jphysiol.2005.093526

[pii]

Hamann, S., Herrera-Perez, J. J., Zeuthen, T., & Alvarez-Leefmans, F. J. (2010).

Cotransport of water by the Na+-K+-2Cl- cotransporter NKCC1 in mammalian

epithelial cells. The Journal of Physiology, 588(Pt 21), 4089-4101. doi:10.1113/jphysiol.2010.194738 [doi]

Hara-Chikuma M, Sohara E, Rai T, Ikawa M, Okabe M, Sasaki S, Uchida S &

Verkman AS 2005 Progressive adipocyte hypertrophy in aquaporin-7-deficient mice: adipocyte glycerol permeability as a novel regulator of fat accumulation. Journal of

Biological Chemistry 280 15493–15496. (doi:10.1074/jbc.C500028200)

Hasegawa, H., Ma, T., Skach, W., Matthay, M. A., & Verkman, A. S. (1994).

Molecular cloning of a mercurial-insensitive water channel expressed in selected water-

transporting tissues. The Journal of Biological Chemistry, 269(8), 5497-5500.

Hermo, L., Schellenberg, M., Liu, L. Y., Dayanandan, B., Zhang, T., Mandato, C.

A., et al. (2008). Membrane domain specificity in the spatial distribution of aquaporins 5,

7, 9, and 11 in efferent ducts and epididymis of rats. The Journal of Histochemistry and

Cytochemistry : Official Journal of the Histochemistry Society, 56(12), 1121-1135. doi:10.1369/jhc.2008.951947 [doi]

Hertz, L., Song, D., Xu, J., Du, T., Gu, L., & Peng, L. (2013). When can AQP4 assist transporter-mediated K(+) uptake? The Journal of General Physiology, 142(1), 87-

89. doi:10.1085/jgp.201310977 [doi]

Hladky, S. B., & Barrand, M. A. (2014). Mechanisms of fluid movement into, through and out of the brain: Evaluation of the evidence. Fluids and Barriers of the CNS,

11(1), 26-8118-11-26. eCollection 2014. doi:10.1186/2045-8118-11-26 [doi]

Hibuse T, Maeda N, Funahashi T, Yamamoto K, Nagasawa A, Mizunoya W,

Kishida K, Inoue K, Kuriyama H, Nakamura T et al. 2005 Aquaporin 7 deficiency is associated with development of obesity through activation of adipose glycerol kinase.

PNAS 102 10993–10998.(doi:10.1073/pnas.0503291102)

Huang, Y., Tracy, R., Walsberg, G. E., Makkinje, A., Fang, P., Brown, D., et al.

(2001). Absence of aquaporin-4 water channels from kidneys of the desert rodent dipodomys merriami merriami. American Journal of Physiology.Renal Physiology,

280(5), F794-802.

Husted, R. F., & Reed, D. J. (1976). Regulation of cerebrospinal fluid potassium by the cat choroid plexus. The Journal of Physiology, 259(1), 213-221.

Ishibashi, K., Kuwahara, M., Gu, Y., Kageyama, Y., Tohsaka, A., Suzuki, F., et al. (1997). Cloning and functional expression of a new water channel abundantly expressed in the testis permeable to water, glycerol, and urea. The Journal of Biological

Chemistry, 272(33), 20782-20786.

Ishibashi, K., Kuwahara, M., Gu, Y., Tanaka, Y., Marumo, F., & Sasaki, S.

(1998). Cloning and functional expression of a new aquaporin (AQP9) abundantly expressed in the peripheral leukocytes permeable to water and urea, but not to glycerol.

Biochemical and Biophysical Research Communications, 244(1), 268-274. doi:S0006-

291X(98)98252-3 [pii]

Jelen S, Wacker S, Aponte-Santamaria C, Skott M, Rojek A, Johanson U,

Kjellbom P, Nielsen S, de Groot BL & Rutzler M 2011 Aquaporin-9 protein is the primary route of hepatocyte glycerol uptake for glycerol gluconeogenesis in mice.

Journal of Biological Chemistry 286-44319–44325. (doi:10.1074/jbc.M111.297002)

Jin, B. J., Zhang, H., Binder, D. K., & Verkman, A. S. (2013). Aquaporin-4- dependent K(+) and water transport modeled in brain extracellular space following neuroexcitation. The Journal of General Physiology, 141(1), 119-132. doi:10.1085/jgp.201210883 [doi]

Jimenez, A. J., Dominguez-Pinos, M. D., Guerra, M. M., Fernandez-Llebrez, P.,

& Perez-Figares, J. M. (2014). Structure and function of the ependymal barrier and diseases associated with ependyma disruption. Tissue Barriers, 2, e28426. doi:10.4161/tisb.28426 [doi]

Johanson, C. E., Duncan, J. A.,3rd, Klinge, P. M., Brinker, T., Stopa, E. G., &

Silverberg, G. D. (2008). Multiplicity of cerebrospinal fluid functions: New challenges in health and disease. Cerebrospinal Fluid Research, 5, 10-8454-5-10. doi:10.1186/1743-

8454-5-10 [doi]

Johansson, P. A., Dziegielewska, K. M., Ek, C. J., Habgood, M. D., Mollgard, K.,

Potter, A., et al. (2005). Aquaporin-1 in the choroid plexuses of developing mammalian brain. Cell and Tissue Research, 322(3), 353-364. doi:10.1007/s00441-005-1120-x [doi]

Jung, J. S., Bhat, R. V., Preston, G. M., Guggino, W. B., Baraban, J. M., & Agre,

P. (1994). Molecular characterization of an aquaporin cDNA from brain: Candidate osmoreceptor and regulator of water balance. Proceedings of the National Academy of

Sciences of the United States of America, 91(26), 13052-13056.

Katada, R., Akdemir, G., Asavapanumas, N., Ratelade, J., Zhang, H., & Verkman,

A. S. (2014). Greatly improved survival and neuroprotection in aquaporin-4-knockout mice following global cerebral ischemia. FASEB Journal : Official Publication of the

Federation of American Societies for Experimental Biology, 28(2), 705-714. doi:10.1096/fj.13-231274 [doi]

Keep, R. F., Xiang, J., & Betz, A. L. (1994). Potassium cotransport at the rat choroid plexus. The American Journal of Physiology, 267(6 Pt 1), C1616-22.

Kitada, M., Chakrabortty, S., Matsumoto, N., Taketomi, M., & Ide, C. (2001).

Differentiation of choroid plexus ependymal cells into astrocytes after grafting into the pre-lesioned spinal cord in mice. Glia, 36(3), 364-374. doi:10.1002/glia.1123 [pii]

Krylov, A. V., Pohl, P., Zeidel, M. L., & Hill, W. G. (2001). Water permeability of asymmetric planar lipid bilayers: Leaflets of different composition offer independent and additive resistances to permeation. The Journal of General Physiology, 118(4), 333-

340.

Kuriyama, H., Shimomura, I., Kishida, K., Kondo, H., Furuyama, N., Nishizawa,

H., et al. (2002). Coordinated regulation of fat-specific and liver-specific glycerol channels, aquaporin adipose and aquaporin 9. Diabetes, 51(10), 2915-2921.

Larsen, B. R., Assentoft, M., Cotrina, M. L., Hua, S. Z., Nedergaard, M., Kaila,

K., et al. (2014). Contributions of the na(+)/K(+)-ATPase, NKCC1, and Kir4.1 to hippocampal K(+) clearance and volume responses. Glia, 62(4), 608-622. doi:10.1002/glia.22629 [doi]

Laforenza, U., Miceli, E., Gastaldi, G., Scaffino, M. F., Ventura, U., Fontana, J.

M., et al. (2010). Solute transporters and aquaporins are impaired in celiac disease.

Biology of the Cell / Under the Auspices of the European Cell Biology Organization,

102(8), 457-467. doi:10.1042/BC20100023 [doi]

Laforenza, U., Scaffino, M. F., & Gastaldi, G. (2013). Aquaporin-10 represents an alternative pathway for glycerol efflux from human adipocytes. PloS One, 8(1), e54474. doi:10.1371/journal.pone.0054474 [doi]

Lebeck, J. (2014). Metabolic impact of the glycerol channels AQP7 and AQP9 in adipose tissue and liver. Journal of Molecular Endocrinology, 52(2), R165-78. doi:10.1530/JME-13-0268 [doi]

Lonskaya, I., Hebron, M. L., Algarzae, N. K., Desforges, N., & Moussa, C. E.

(2013). Decreased parkin solubility is associated with impairment of autophagy in the nigrostriatum of sporadic parkinson's disease. Neuroscience, 232, 90-105. doi:10.1016/j.neuroscience.2012.12.018 [doi]

Lui, J. H., Hansen, D. V., & Kriegstein, A. R. (2011). Development and evolution of the human neocortex. Cell, 146(1), 18-36. doi:10.1016/j.cell.2011.06.030 [doi]

Macaulay, N., & Zeuthen, T. (2012). Glial K(+) clearance and cell swelling: Key roles for cotransporters and pumps. Neurochemical Research, 37(11), 2299-2309. doi:10.1007/s11064-012-0731-3 [doi]

Maeda N, Funahashi T & Shimomura I (2008) Metabolic impact of adipose and hepatic glycerol channels aquaporin 7 and aquaporin 9. Nature Clinical Practice.

Endocrinology & Metabolism 4 627–634.(doi:10.1038/ncpendmet0980)

Maeda N, Hibuse T & Funahashi T (2009) Role of aquaporin-7 and aquaporin-9

in glycerol metabolism; involvement in obesity. Handbook of Experimental

Pharmacology 190 233–249. (doi:10.1007/978-3-540-79885-9_12)

Milhorat, T. H., Hammock, M. K., Fenstermacher, J. D., & Levin, V. A. (1971).

Cerebrospinal fluid production by the choroid plexus and brain. Science (New York,

N.Y.), 173(3994), 330-332.

Miller, E. W., Dickinson, B. C., & Chang, C. J. (2010). Aquaporin-3 mediates hydrogen peroxide uptake to regulate downstream intracellular signaling. Proceedings of the National Academy of Sciences of the United States of America, 107(36), 15681-

15686. doi:10.1073/pnas.1005776107 [doi]

Mobasheri, A., Marples, D., Young, I. S., Floyd, R. V., Moskaluk, C. A., &

Frigeri, A. (2007). Distribution of the AQP4 water channel in normal human tissues:

Protein and tissue microarrays reveal expression in several new anatomical locations, including the prostate gland and seminal vesicles. Channels (Austin, Tex.), 1(1), 29-38.

Musa-Aziz, R., Chen, L. M., Pelletier, M. F., & Boron, W. F. (2009). Relative

CO2/NH3 selectivities of AQP1, AQP4, AQP5, AmtB, and RhAG. Proceedings of the

National Academy of Sciences of the United States of America, 106(13), 5406-5411. doi:10.1073/pnas.0813231106 [doi]

Mylonakou, M. N., Petersen, P. H., Rinvik, E., Rojek, A., Valdimarsdottir, E.,

Zelenin, S., et al. (2009). Analysis of mice with targeted deletion of AQP9 gene provides conclusive evidence for expression of AQP9 in neurons. Journal of Neuroscience

Research, 87(6), 1310-1322. doi:10.1002/jnr.21952 [doi]

Nagelhus, E. A., Mathiisen, T. M., & Ottersen, O. P. (2004). Aquaporin-4 in the central nervous system: Cellular and subcellular distribution and coexpression with

KIR4.1. Neuroscience, 129(4), 905-913. doi:S0306-4522(04)00755-9 [pii]

Nagelhus, E. A., & Ottersen, O. P. (2013). Physiological roles of aquaporin-4 in brain. Physiological Reviews, 93(4), 1543-1562. doi:10.1152/physrev.00011.2013 [doi]

Nazari, Z., Nabiuni, M., Safaei Nejad, Z., Delfan, B., & Irian, S. (2015).

Expression of aquaporins in the rat choroid plexus. Arch Neurosci, 2(1 SP e17312)

Nejsum, L. N., Elkjaer, M., Hager, H., Frokiaer, J., Kwon, T. H., & Nielsen, S.

(2000). Localization of aquaporin-7 in rat and mouse kidney using RT-PCR, immunoblotting, and immunocytochemistry. Biochemical and Biophysical Research

Communications, 277(1), 164-170. doi:10.1006/bbrc.2000.3638 [doi]

Nicchia, G. P., Frigeri, A., Nico, B., Ribatti, D., & Svelto, M. (2001). Tissue distribution and membrane localization of aquaporin-9 water channel: Evidence for sex- linked differences in liver. The Journal of Histochemistry and Cytochemistry : Official

Journal of the Histochemistry Society, 49(12), 1547-1556.

Nielsen, S., Nagelhus, E. A., Amiry-Moghaddam, M., Bourque, C., Agre, P., &

Ottersen, O. P. (1997). Specialized membrane domains for water transport in glial cells:

High-resolution immunogold cytochemistry of aquaporin-4 in rat brain. The Journal of

Neuroscience : The Official Journal of the Society for Neuroscience, 17(1), 171-180.

Nielsen, S., Smith, B. L., Christensen, E. I., Knepper, M. A., & Agre, P. (1993).

CHIP28 water channels are localized in constitutively water-permeable segments of the nephron. The Journal of Cell Biology, 120(2), 371-383.

Oshio, K., Binder, D. K., Yang, B., Schecter, S., Verkman, A. S., & Manley, G.

T. (2004). Expression of aquaporin water channels in mouse spinal cord. Neuroscience,

127(3), 685-693. doi:10.1016/j.neuroscience.2004.03.016 [doi]

Oshio, K., Song, Y., Verkman, A. S., & Manley, G. T. (2003). Aquaporin-1 deletion reduces osmotic water permeability and cerebrospinal fluid production. Acta

Neurochirurgica.Supplement, 86, 525-528.

Oshio, K., Watanabe, H., Song, Y., Verkman, A. S., & Manley, G. T. (2005).

Reduced cerebrospinal fluid production and intracranial pressure in mice lacking choroid plexus water channel aquaporin-1. FASEB Journal : Official Publication of the

Federation of American Societies for Experimental Biology, 19(1), 76-78. doi:04-1711fje

[pii]

Papadopoulos, M. C., & Verkman, A. S. (2013). Aquaporin water channels in the

nervous system. Nature Reviews.Neuroscience, 14(4), 265-277. doi:10.1038/nrn3468

[doi]

Paulson, O. B., & Newman, E. A. (1987). Does the release of potassium from

astrocyte endfeet regulate cerebral blood flow? Science (New York, N.Y.), 237(4817),

896-898.

Peters A. & Swan R.C. (1979) The choroid plexus of the mature and aging rat: the choroidal epithelium, Anat. Rec. 194979.325–345

Piechotta, K., Lu, J., & Delpire, E. (2002). Cation chloride cotransporters interact with the stress-related kinases Ste20-related proline-alanine-rich kinase (SPAK) and oxidative stress response 1 (OSR1). The Journal of Biological Chemistry, 277(52),

50812-50819. doi:10.1074/jbc.M208108200 [doi]

Plotkin, M. D., Snyder, E. Y., Hebert, S. C., & Delpire, E. (1997). Expression of the na-K-2Cl cotransporter is developmentally regulated in postnatal rat brains: A possible mechanism underlying GABA's excitatory role in immature brain. Journal of

Neurobiology, 33(6), 781-795. doi:10.1002/(SICI)1097-4695(19971120)33:6<781::AID-

NEU6>3.0.CO;2-5 [pii]

Pollay, M., & Curl, F. (1967). Secretion of cerebrospinal fluid by the ventricular ependyma of the rabbit. The American Journal of Physiology, 213(4), 1031-1038.

Praetorius, J., & Nielsen, S. (2006). Distribution of sodium transporters and

aquaporin-1 in the human choroid plexus. American Journal of Physiology.Cell

Physiology, 291(1), C59-67. doi:00433.2005 [pii]

Prescott L. and Brightman MW . Normal and Pathologically Altered Structures of

the Choroid Plexus. Chapter 2 in: The Blood–Cerebrospinal Fluid Barrier Zheng W and

Chodobski, A. Eds pp.25-52, Chapman & Hall/CRC, Taylor & Francis Group. Boca

Raton, Fl 2005.

Preston, G. M., Carroll, T. P., Guggino, W. B. and Agre, P.(1992). Appearance of water channels in Xenopus oocytes expressing red cell CHIP28 protein. Science 256,

385-387

Rash, J. E., Yasumura, T., Hudson, C. S., Agre, P., & Nielsen, S. (1998). Direct immunogold labeling of aquaporin-4 in square arrays of astrocyte and ependymocyte plasma membranes in rat brain and spinal cord. Proceedings of the National Academy of

Sciences of the United States of America, 95(20), 11981-11986.

Roales-Bujan, R., Paez, P., Guerra, M., Rodriguez, S., Vio, K., Ho-Plagaro, A., et al. (2012). Astrocytes acquire morphological and functional characteristics of ependymal cells following disruption of ependyma in hydrocephalus. Acta Neuropathologica,

124(4), 531-546. doi:10.1007/s00401-012-0992-6 [doi]

Rodriguez A, Catalan V, Gomez-Ambrosi J, Garcia-Navarro S, Rotellar F,Valenti

V, Silva C, Gil MJ, Salvador J, Burrell MA et al. 2011 Insulin- and leptin-mediated

control of aquaglyceroporins in human adipocytes and hepatocytes is mediated via the

PI3K/Akt/mTOR signaling cascade. Journal of Clinical Endocrinology and Metabolism

96 E586–E597.(doi:10.1210/jc.2010-1408)

Rojek, A. M., Skowronski, M. T., Fuchtbauer, E. M., Fuchtbauer, A. C., Fenton,

R. A., Agre, P., et al. (2007). Defective glycerol metabolism in aquaporin 9 (AQP9) knockout mice. Proceedings of the National Academy of Sciences of the United States of

America, 104(9), 3609-3614. doi:0610894104 [pii]

Rosenthal, R., Milatz, S., Krug, S. M., Oelrich, B., Schulzke, J. D., Amasheh, S., et al. (2010). Claudin-2, a component of the tight junction, forms a paracellular water channel. Journal of Cell Science, 123(Pt 11), 1913-1921. doi:10.1242/jcs.060665 [doi]

Sakai, H., Sato, K., Kai, Y., Shoji, T., Hasegawa, S., Nishizaki, M., et al. (2014).

Distribution of aquaporin genes and selection of individual reference genes for

quantitative real-time RT-PCR analysis in multiple tissues of the mouse. Canadian

Journal of Physiology and Pharmacology, 92(9), 789-796. doi:10.1139/cjpp-2014-0157

[doi]

Segal, M. B. (1993). Extracellular and cerebrospinal fluids. Journal of Inherited

Metabolic Disease, 16(4), 617-638.

Shin, I., Kim, H. J., Lee, J. E., & Gye, M. C. (2006). Aquaporin7 expression during perinatal development of mouse brain. Neuroscience Letters, 409(2), 106-111. doi:S0304-3940(06)00996-7 [pii]

Smith, D. E., Johanson, C. E., & Keep, R. F. (2004). Peptide and peptide analog transport systems at the blood-CSF barrier. Advanced Drug Delivery Reviews, 56(12),

1765-1791. doi:10.1016/j.addr.2004.07.008 [doi]

Sohara, E., Ueda, O., Tachibe, T., Hani, T., Jishage, K., Rai, T., et al. (2007).

Morphologic and functional analysis of sperm and testes in aquaporin 7 knockout mice.

Fertility and Sterility, 87(3), 671-676. doi:S0015-0282(06)04008-8 [pii]

Spassky, N., Merkle, F. T., Flames, N., Tramontin, A. D., Garcia-Verdugo, J. M.,

& Alvarez-Buylla, A. (2005). Adult ependymal cells are postmitotic and are derived from radial glial cells during embryogenesis. The Journal of Neuroscience : The Official

Journal of the Society for Neuroscience, 25(1), 10-18. doi:25/1/10 [pii]

Speake, T., Freeman, L. J., & Brown, P. D. (2003). Expression of aquaporin 1 and aquaporin 4 water channels in rat choroid plexus. Biochimica Et Biophysica Acta,

1609(1), 80-86. doi:S0005273602006582 [pii]

Spector, R., & Johanson, C. E. (1989). The mammalian choroid plexus. Scientific

American, 261(5), 68-74.

Suzuki-Toyota, F., Ishibashi, K., & Yuasa, S. (1999). Immunohistochemical

localization of a water channel, aquaporin 7 (AQP7), in the rat testis. Cell and Tissue

Research, 295(2), 279-285.

Takano, T., Tian, G. F., Peng, W., Lou, N., Libionka, W., Han, X., et al. (2006).

Astrocyte-mediated control of cerebral blood flow. Nature Neuroscience, 9(2), 260-267.

doi:nn1623 [pii]

Terris, J., Ecelbarger, C. A., Marples, D., Knepper, M. A., & Nielsen, S. (1995).

Distribution of aquaporin-4 water channel expression within rat kidney. The American

Journal of Physiology, 269(6 Pt 2), F775-85.

Thomas, D., Bron, P., Ranchy, G., Duchesne, L., Cavalier, A., Rolland, J. P., et al.

(2002). Aquaglyceroporins, one channel for two molecules. Biochimica Et Biophysica

Acta, 1555(1-3), 181-186. doi:S000527280200275X [pii]

Van Hoek, A. N., Ma, T., Yang, B., Verkman, A. S., & Brown, D. (2000).

Aquaporin-4 is expressed in basolateral membranes of proximal tubule S3 segments in mouse kidney. American Journal of Physiology.Renal Physiology, 278(2), F310-6.

Venero, J. L., Vizuete, M. L., Ilundain, A. A., Machado, A., Echevarria, M., &

Cano, J. (1999). Detailed localization of aquaporin-4 messenger RNA in the CNS:

Preferential expression in periventricular organs. Neuroscience, 94(1), 239-250. doi:S0306-4522(99)00182-7 [pii]

Verkman, A. S. (2011). Aquaporins at a glance. Journal of Cell Science, 124(Pt

13), 2107-2112. doi:10.1242/jcs.079467 [doi]

Verkman, A. S., & Mitra, A. K. (2000). Structure and function of aquaporin water

channels. American Journal of Physiology.Renal Physiology, 278(1), F13-28.

Verkman, A. S. (2013). Aquaporins. Current Biology, 23(2), R52-R55. Retrieved

from http://www.sciencedirect.com/science/article/pii/S0960982212013693

Wahlby, C., Erlandsson, F., Bengtsson, E., & Zetterberg, A. (2002). Sequential

immunofluorescence staining and image analysis for detection of large numbers of antigens in individual cell nuclei. Cytometry, 47(1), 32-41. doi:10.1002/cyto.10026 [pii]

Wakayama, Y., Inoue, M., Kojima, H., Jimi, T., Shibuya, S., Hara, H., et al.

(2004). Expression and localization of aquaporin 7 in normal skeletal myofiber. Cell and

Tissue Research, 316(1), 123-129. doi:10.1007/s00441-004-0857-y [doi]

Walz, T., Hirai, T., Murata, K., Heymann, J. B., Mitsuoka, K., Fujiyoshi, Y., et al.

(1997). The three-dimensional structure of aquaporin-1. Nature, 387(6633), 624-627. doi:10.1038/42512 [doi]

Wu, Q., Delpire, E., Hebert, S. C., & Strange, K. (1998). Functional demonstration of Na+-K+-2Cl- cotransporter activity in isolated, polarized choroid plexus

cells. The American Journal of Physiology, 275(6 Pt 1), C1565-72.

Xu, G. Y., Wang, F., Jiang, X., & Tao, J. (2010). Aquaporin 1, a potential

therapeutic target for migraine with aura. Molecular Pain, 6, 68-8069-6-68. doi:10.1186/1744-8069-6-68 [doi]

Yan, Y., Dempsey, R. J., & Sun, D. (2001). Expression of Na+-K+-Cl- cotransporter in rat brain during development and its localization in mature astrocytes.

Brain Research, 911(1), 43-55. doi:S0006-8993(01)02649-X [pii]

Yan, Y., Dempsey, R. J., & Sun, D. (2001). Na+-K+-Cl- cotransporter in rat focal

cerebral ischemia. Journal of Cerebral Blood Flow and Metabolism: Official Journal of

the International Society of Cerebral Blood Flow and Metabolism, 21(6), 711-721.

doi:10.1097/00004647-200106000-00009 [doi]

Yang, B., Brown, D., & Verkman, A. S. (1996). The mercurial insensitive water

channel (AQP-4) forms orthogonal arrays in stably transfected chinese hamster ovary

cells. The Journal of Biological Chemistry, 271(9), 4577-4580.

Yang, B., Verbavatz, J. M., Song, Y., Vetrivel, L., Manley, G., Kao, W. M., et al.

(2000). Skeletal muscle function and water permeability in aquaporin-4 deficient mice.

American Journal of Physiology.Cell Physiology, 278(6), C1108-15.

Zeuthen, T. (1994). Cotransport of K+, Cl- and H2O by membrane proteins from

choroid plexus epithelium of necturus maculosus. The Journal of Physiology, 478 ( Pt

2)(Pt 2), 203-219.

Zeuthen, T., & Macaulay, N. (2012). Cotransport of water by Na+-K+-Cl-

cotransporters expressed in xenopus oocytes: NKCC1 versus NKCC2. The Journal of

Physiology, 590(Pt 5), 1139-1154. doi:10.1113/jphysiol.2011.226316 [doi]