Impact of the deletion of syntrophins and AQP4 on the prevalence of astrocytic gap junctions

(1)

Impact of the deletion of syntrophins and AQP4 on the prevalence of

astrocytic gap junctions

Nenad Jovanovic

Master Thesis for the title of Master in Pharmacy Department of Pharmaceutical Biosciences

School of Pharmacy 45 credits

The Faculty of Mathematics and Natural Sciences UNIVERSITETET I OSLO

April 2020

(2)

II

(3)

III

Impact of deletion of syntrophins and AQP4 on the prevalence of astrocytic gap

junctions

(4)

IV

Impact of deletion of syntrophins and AQP4 on the prevalence of astrocytic gap junctions Nenad Jovanovic

http://www.duo.uio.no/

Trykk: Reprosentralen, Universitetet i Oslo

(5)

V

Acknowledgements

This project was conducted at the Department of Anatomy, Institute of Basic Medical

Sciences, University of Oslo, between September 2019 and April 2020, under the supervision of Professor Mahmood Amiry-Moghaddam and Ph.D. student Nadia Skauli.

I feel privileged to have been involved in the work of this research group. Many thanks to my mentors for their support and advice while working on this thesis. I owe a special gratitude to my supervisor professor Mahmood Amiry-Moghaddam for identifying practical areas of research I was particularly interested in, and for creating a tailor-made project. On the other hand, Nadia Skauli had unlimited patience for me and helped me a lot with her tips and feedback.

Many thanks to Senior Engineer Bashir Ahmad Hakim and Principal Engineer Rao Shreyas for help and support, as well as others in the research group.

Professor Cecilie Morland, thank you for your kind help and encouragement.

Also a big thanks to everyone who supported me on this topic - my family, friends, and special thanks to my dear colleague Knut Breistøl.

Tønsberg April 2020.

Nenad Jovanovic

(6)

VI

(7)

VII

Abstract

Astrocytes are the most prevalent glial cells in the brain, with foot processes surrounding microvessels and neuronal compartments. Astrocytes are important players in maintenance of water and ion homeostasis, a key feature critical for normal brain function. Two classes of proteins are essential in the homeostatic functions of astrocytes: 1) aquaporin-4 (AQP4) water channels mainly localized at the astrocytic foot processes surrounding the brain microvessels, and 2) astrocytic connexins (Connexin-43 and Connexin-30) forming gap junctions (GJ) between neighboring astrocytes. This master project is based on indications of the

interconnection and reciprocal regulation between AQP4 and astrocytic gap junctions in the brain.

This study is a continuation of recent findings in the host laboratory showing an increase in the number of astrocytic gap junctions in the hippocampus of mice with targeted deletion of AQP4. The aim was to investigate whether it is the AQP4’s endfeet localization or its expression level which is responsible for the observed increase in the number of astrocytic gap junctions. We therefore studied two models: 1) mice with heterozygous (Het) and homozygous deletion of AQP4 where AQP4 expression is halved (Het) or completely lost – knockout (KO) at protein and mRNA level, and 2) mice with genetic deletion of syntrophins, where AQP4 is mislocalized from the perivascular membranes, without affecting protein and mRNA expression levels.

Post-embedding immunogold electron microscopy, using specific antibodies to Connexin 43 (Cx43), was used as the method to quantify gap junctions in the mouse models described above.

A significant increase in the number of gap junctions in mice with concomitant deletion of α- and β-syntrophin (αβ-syntrophin KO) was demonstrated. On the other hand, we failed to demonstrate any effect of a heterozygous or homozygous deletion of AQP4 on the number of gap junctions.

(8)

VIII AQP4, as well as Connexin 43, as a structural part of gap junctions in astrocytes, are an attractive target for the treatment of a variety of diseases and disorders. However, more research is needed to understand all aspects of their functions and effects on physiological and pathological processes, as well as their interrelationships, which is what this master project deals with.

(9)

IX

(10)

X

1 Introduction

In this project, I have studied the possible connection between water channels, their anchoring proteins and the expression of astrocytic gap junctions. Here, I will introduce the cell types we are studying, our main proteins of interest and related diseases and disorders.

1.1 Types of cells in the Central Nervous system (CNS)

1.1.1 Glial cells

The brain parenchyma mainly consists of two types of cells – neurons and glial cells.

Neurons are cells responsible for sending and receiving of signals and information in the nervous system.

Glial cells have different functions depending on the type. There are three types of glial cells:

oligodendrocytes, microglia and astrocytes.

Oligodendrocytes are associated with production of the myelin, while microglial cells have a role in the immune system of the CNS.

Astrocytes are the main focus of this thesis and their functions will be addressed in more detail.

In addition to neurons and glial cells, other important cell types are part of the CNS. Pericytes are cells present at intervals along the capillary walls. They are important for formation of the blood vessel and maintenance of the blood–brain barrier (BBB) (Attwell et al. 2016).

Endothelial cells make up the walls of the blood vessels. Brain microvascular endothelial cells are located between nervous tissue and circulating blood, and have important role in regulating CNS homeostasis (Yu et al. 2015).

(13)

2 Figure 1 – Astrocytes and endothelial cells in the brain (Iadecola and Nedergaard 2007).

AQP4 are located in astrocyte endfeet. Gap junctions can be found between neighboring astrocytes. Pericytes are present at intervals along the capillary walls. Other protein structures also exist in the astrocyte endfeet such as glucose transporters and K⁺ channels.

(14)

3 1.1.2 Astrocyte function

Glial cells were identified and for the first time described by the French physician René Dutrochet in 1824. However, we had to wait until 1856 when the German pathologist Rudolph Virchow came up with the term “nervekitt,” (nerve glue) to describe what was considered a kind of connective tissue of the CNS (Gentile and Colucci D'Amato 2018). The cellular nature of this tissue was soon recognized and different types of glial cells were morphologically characterized (Parpura and Verkhratsky 2012). Astrocytes were mainly considered to be supporting cells in the brain involved in structural and functional support to neurons (Nag 2011; Kimelberg and Nedergaard 2010).

The name astrocyte (astro meaning “star”) was first used by Michael von Lenhossek in 1891.

The name comes from the astrocytic processes that give the cell shape of a star. Astrocytes have processes abutting brain microvessels and the pial surface called endfeet, and in addition also processes surrounding the neuronal compartments (Parpura and Verkhratsky 2012). The pia is a delicate, highly vascular layer covering the brain and spinal cord (Adeeb et al. 2013).

Today we are familiar with different functions of astrocytes in the CNS, such as glutamate, ions (i.e., Ca2+, K+) and water homeostasis, also blood-brain barrier functions (Verkhratsky and Nedergaard 2018). Astrocyte express a large number of membrane channels including aquaporin water channels, neurotransporter uptake proteins and ion channels, including voltage-gated Ca2+ channels and K+ channels (Kim, Park, and Choi 2019). They are also tightly connected with neuronal functions and synapse modulation, and also have active role in response to CNS injury (Kim, Park, and Choi 2019; Santello, Toni, and Volterra 2019).

The role and function of astrocytes is diverse and has been the subject of much research partly because of their potential as a therapeutic targets in relation to a number of neuropathological conditions, such as Alzheimer’s disease, Parkinson’s disease, and stroke (Kim, Park, and Choi 2019).

(15)

4 1.1.3 The Blood-Brain Barrier (BBB)

The blood-brain barrier (BBB) is a diffusion barrier, which controls transport of most compounds from blood to brain. BBB consist of three elements - endothelial cells, astrocyte end-feet, and pericytes (cells that wrap around endothelial cells throughout the body). Tight junctions, located between the cerebral endothelial cells, form a diffusion barrier (Castro Dias et al. 2019).

Vascular endothelial cells and pericytes are surrounded by a layer of basal lamina. Basal lamina is made up of extracellular proteins secreted by different vascular and neural cells.

BBB has significant role in regulating the movement of ions, molecules, and cells between the blood and the central nervous system. This barrier provides the suitable environment for proper neural function, also protects the CNS from injury and disease. Dysfunction of the BBB can lead to a number of neurological diseases and disorders including stroke (Daneman 2012; Ballabh, Braun, and Nedergaard 2004). Astrocytic endfeet enwrap almost completely the perivascular basal lamina and is considered as part of the BBB (Mathiisen et al. 2010).

Figure 2 - The blood-brain barrier (van de Haar et al. 2015). Endothelial cells are marked in purple, pericytes in turquoise, astrocyte endfeet in green.

This work focuses on important protein molecules involved in water, ion and small molecule transport in the brain, localized in the astrocytic endfeet, namely the water channel

aquaporin-4 (AQP4) and the gap junction protein Connexin 43 (Cx43).

(16)

5 1.1.4 Regions in the human brain

Human brain consists of three main parts:

1. The Cerebrum is the largest part of the brain. It is composed of the right and left hemisphere and it is in charge of advanced brain functions such as learning, emotions etc.

2. The Cerebellum which is located under the cerebrum. Its function is among others to coordinate muscle movements.

3. The Brainstem connecting the cerebrum and cerebellum to the spinal cord. It carries out many automatic functions such as breathing and beating of the heart (Ackerman 1992).

Figure 3 – brain regions and subregions (Ackerman 1992). Cerebral cortex occupies the greatest surface area of the human brain. Cerebellum are located under Cerebrum, also brainstem connecting it with the spinal cord.

(17)

6 The hippocampus is a complex brain structure that lies sub-cortically and have important role in learning and memory (Anand and Dhikav 2012). In this study, the hippocampus was a region of interest. This is based on findings that the increase in connexin gap junction

formation in AQP4 KO mice was more prominent in hippocampus than in parietal cortex (Katoozi et al. 2017). Previous studies have also reported regional differences in the regulation of expression astrocytic gap junctions (Ball et al. 2011; Aberg et al. 2000). The hippocampus is especially suitable for research on mouse models because of its similar neuroanatomy and function with the human brain (Clark and Squire 2013). Furthermore, we study the hippocampus because it is often damaged by disease. The hippocampus is often affected in neuropathologies such as Alzheimer’s disease and stroke (Halliday 2017; Walha et al. 2013), conditions where astrocytes play an important role in the disease process.

1.2 Aquaporin-4 (AQP4) – the main water channel in the brain

Aquaporins are protein molecules which have a significant role in water transport. Water movement crosswise cell membranes is facilitated by these pores. The first aquaporin was discovered by Peter Agre (Agre 2004; Preston et al. 1992). Agre received the Nobel Prize in Chemistry in 2003 for this finding. There are currently 13 known aquaporins in mammals, involved in many physiological processes such as urinary concentration, body fluid

homeostasis, brain function and many more (Brown 2017).

AQP4 is the most abundant aquaporin in the brain and is mainly localized at the perivascular astrocyte endfeet abutting brain microvessels (Rash et al. 1998; Nielsen et al. 1997). The perivascular pool of AQP4 serves as both an entry route as well as an exit route for water molecules (Amiry-Moghaddam, Otsuka, et al. 2003).

(18)

7 Figure 4 - AQP4 structure (Iacovetta, Rudloff, and Kirby 2012). The molecule extends

through the cell membrane 6 times, arranging 5 loops classified as A, C, and E on the extracellular side and B and D on the intracellular. The aspargine‐proline‐alanine (NPA) segments in loops B and E are in charge of selectivity of the channels. The AQP4 protein is folded in a central hour‐glass shape formed by the N-1-naphthylphthalamic acid (NPA) motif which only allows for passage of water molecules.

AQP4 is involved in the regulation of ion and water homeostasis (Amiry-Moghaddam and Ottersen 2003). AQP4 is also involved in the development of brain edema and furthermore associated with different disease processes in the brain, such as seizures and tumors. In the course of brain edema, astrocytes are the major cell type that swell due to water influx. The influx occurs primarily at the astrocyte endfeet where AQP4 is mainly located. Manley et al.

documented a connection between AQP4 and brain edema, where knocking out AQP4 led to reduced extent of edema after experimental stroke and hypo-osmotic stress (Manley et al.

2000).

We investigate aquaporins in this project because AQP4 is an attractive pharmacological target and consequential significance of understanding the regulations of AQP4.

(19)

8

1.3 Syntrophins – the anchoring molecules of AQP4

Syntrophins represents a family of various cytoplasmic membrane-associated adaptor

proteins. They appear for an important element of many signaling events, having a function in various signaling complexes (Bhat, Adams, and Khanday 2013).

The syntrophin protein family is composed of five known homologous isoforms: alpha-1- (α1-) syntrophin, beta-1(ß1-) syntrophin, beta-2- (ß2-) syntrophin, gamma-1(γ1-) syntrophin, and gamma-2- (γ2-) syntrophin (Ahn and Kunkel 1993; Ahn et al. 1996; Adams et al. 1993).

Being a member of the dystrophin complex α-syntrophin is responsible for anchoring of AQP4 in the brain (Amiry-Moghaddam, Frydenlund, and Ottersen 2004). Amiry-Moghaddam et al. (2004) used mice deficient in α-syntrophin where the perivascular pool of AQP4 were removed for assessment of its functional roles. Deletion of α-syntrophin caused however a mislocalization rather than a loss of AQP4, by other means redistribution of AQP4 from endfeet to non-endfeet membranes, without affecting its total level.

Figure 5 - Immunogold analyses reveal two pools of AQP4 at the brain–blood interface. The major pool resides in the perivascular membranes of astrocytes (double arrows) and is depleted after α-syntrophin knockout. The minor pool resides in the abluminal (arrows) and adluminal membrane (arrowheads) of endothelial cells and is independent of α-syntrophin (B). (A) Wild type animals; (B) α-syn knockout animals; E, endothelial cell; L, capillary lumen; P, pericyte. (Amiry-Moghaddam, Frydenlund, and Ottersen 2004)

(20)

9 The role of β-syntrophin in the brain is not well known. A recent study from the research group in the host laboratory has elucidated a role of β1-syntrophin in the anchoring of AQP4 in retinal Müller cells. Müller cells are macroglial cells present in retina. This study has shown that deletion of β1-syntrophin causes a partial loss of AQP4 from perivascular retinal Müller cell endfeet. There has also been suggested that upregulation of perivascular α1- syntrophin compensates effect of β1-syntrophin deletion in retina (Katoozi, Rao, et al., 2020), however the roles of β-syntrophin remain to be elucidated.

Deletion of α-syntrophin has previously been shown to lead to a loss of AQP4 from astrocyte membranes surrounding blood vessels (Amiry-Moghaddam, Frydenlund, and Ottersen 2004).

In this study we have used different combinations of syntrophin gene knockout mouse

models. Analysis of α-syntrophin KO mice was part of another ongoing project and therefore this study is focused on β1 and αβ KO which are believed to lead to a similar AQP4

mislocalization as in α-syntrophin KO.

1.3.1 Gap junctions

Proteins known as connexins form so called ‘gap junctions’ between astrocyte perivascular endfeet. Gap junctions consist of two hemichannels of connexin subunits placed in the opposed plasma membranes (Rackauskas, Neverauskas, and Skeberdis 2010).

Figure 6 - Gap junction structure (Spray and Scemes 2013) showing channels consisting of paired connexons (connexin hemichannels) and their organization in the membrane. The six connexin hexamer is a connexon. Gap junction is composed of 12 connexin proteins (two connexons).

(21)

10 These intercellular channels permits the transit of small molecules like ions, second

messengers, amino acids, glucose and molecules in size up to ∼1 kDa in molecular mass (Yeager and Nicholson 1996; Saez et al. 2003). Three separate connexin types are known in astroglial cells of the mouse brain – Cx43, Cx30 and Cx26 (Theis and Steinhäuser 2013). Astrocytic GJ are mainly composed of Cx43 and Cx30 (Nagy et al. 1999; Rouach et al.

2002; Yamamoto et al. 1990; Yamamoto et al. 1992).

Figure 7 - AQP4 channels and connexin gap junctions (Iacovetta, Rudloff, and Kirby 2012).

AQP4 channels in astrocytic endfeet are shown as dots while gap junctions between neighboring astrocytes are shown as lines.

Astocytic gap junctions could be involved in controlling neuronal activity by having a part in controlling clearance of extracellular potassium and glutamate (Pannasch et al. 2011; Wallraff et al. 2006). Gap junctions in astrocytes are also shown to be involved in calcium signaling as a part of Ca ²⁺signaling network (Jorgensen et al. 1997).

Similarly to AQP4, Cx43 is known to be involved in neuropathologies such as stroke, edema, epilepsy and Alzheimer’s disease (Yamasaki 2018; Zador et al. 2009; Boison and Steinhäuser 2018).

(22)

11

1.4 Potential reciprocal regulation of AQP4 and Gap junctions

In AQP4 knockout mice, the number of gap junctions in hippocampus and cortex have been shown to increase compared to wild type (WT) mice (Katoozi et al. 2017). Furthermore, a recent study from the research group in the host laboratory has shown that in Cx43/30 double knockout animals (dKO), total AQP4 expression is decreased (Katoozi, Skauli, et al., 2020).

These findings indicate that AQP4 and gap junctions are reciprocally regulated in the mouse brain.

The mechanism by which deletion of AQP4 leads to the observed changes in gap junction numbers is not known. It is uncertain whether is it the expression level of AQP4 or loss of its domain specific localization in the astrocyte membrane which leads to that.

Reciprocal regulation of AQP4 and Cx43 could be explaining why AQP4 deletion has limited effects on brain volume and K + homeostasis (Katoozi et al. 2017).

It was also demonstrated that mice lacking gap junction proteins Cx30 and Cx43 manifest partial loss of the AQP4 in astrocyte endfeet (Ezan et al. 2012). This also supports the reciprocal theory.

1.4.1 The role of AQP4 and gap junctions in disease There are several diseases where the expression levels and localization of AQP4, and astrocytic connexins/gap junctions have been demonstrated included.

Stroke, edema, epilepsy and neurodegenerative disorders are diseases in which both AQP4 and Cx43 are involved (Katoozi et al. 2020; Amiry-Moghaddam, Williamson, et al. 2003;

Zador et al. 2009; Yamasaki 2018; Boison and Steinhäuser 2018; Lan et al. 2017). Here, reciprocal regulation and understanding of this can be of great importance. Understanding the factors regulating the polarized expression of AQP4 is essential in developing pharmaceutical interventions in conditions where the polarized expression is lost.

(23)

12 AQP4 has been considered to be associated with the removal of interstitial β-amyloid, protein molecule connected with pathogenesis of the Alzheimer disease (Lan et al. 2017). Preventing or reducing amyloid-β accumulation could be an important therapeutic approach for

Alzheimer’s disease (Xu et al. 2015). In Alzheimer's disease both Connexin 43 and gap junction expression in the patient brain samples were shown increased (Yamasaki 2018).

These findings could be indicating important role of connexins in the pathogenesis of the Alzheimer's disease.

In the course of cerebral edema development, AQP4 has been show to advance astrocyte swelling (“cytotoxic swelling”). AQP4 has also been seen to be in charge of reabsorption of extracellular edema fluid (“vasogenic edema”) (Zador et al. 2009). Cytotoxic edema is related to intracellular water accumulation while vasogenic edema with extracellular water

accumulation due to cellular injury or BBB breakdown (Klatzo 1967; Unterberg et al. 2004).

This leads to the conclusion that there is certain potential of AQP4 as a possible therapeutic target in edema associated with stroke (Zador et al. 2009). Agents that block gap junctions appear to be neuroprotective and lead to reduction or deletion of clinical signs of ischemic brain injury (Contreras et al. 2004; Yamasaki 2018).

There are available data that demonstrates an upregulation of AQP4 which occurs at the site of traumatic brain injury in rat models. These data shows also downregulation in adjacent regions and can be considered as a basis for the development of treatments for cerebral edema that accompanies head injury (Sun et al. 2003).

AQP4 expressed in astrocytes has important role in water and K+ homeostasis during neural activity. Because osmolarity and K+ have great effect on epileptic seizures, AQP4 may represent therapeutic target for control of seizures (Binder, Nagelhus, and Ottersen 2012).

Amiry‐Moghaddam et al., (2003) were the first to find a possible connection between syntrophin/AQP4 and epilepsy by demonstrating that deletion of α-syntrophin leads to

delaying of the K+ clearance after neuronal activation, and intensification of epileptic seizures (Amiry-Moghaddam, Williamson, et al. 2003). There are increasing evidences suggesting that gap junctions between astrocytes could be involved in the pathogenesis of epilepsy (Boison and Steinhäuser 2018; Mylvaganam et al. 2014). Several studies have found Cx43 up- regulated with the development of epilepsy (Frederic et al. 2006; Garbelli et al. 2011).

Astrocytic GJ could be possible target in treatment of epilepsy (Li et al. 2019; Mylvaganam et al. 2014).

(24)

13 Neuromyelitis Optica (NMO) is an autoimmune demyelinating disease, which is

characterized by the manifestation of autoantibody (NMO-IgG) to AQP4 (Pisani et al. 2013).

NMO-IgG binding to AQP4 leads to astrocyte injury and Blood Brain Barrier breakdown (Hinson et al. 2012; Hinson et al. 2007). NMO affects the spinal cord and optic nerves (Pache et al. 2017).

To date, few agents targeting AQP4 and Connexin 43 have been developed, but they represent enormous potential for future research. There are many challenges in the development of AQP4 drugs as they show many homologous AQP isoforms, wide distribution in tissues and function which can give possible side effects, also need for high blood-brain barrier

permeation. Despite considerable effort, small molecules targeting AQP4 have not been developed (Verkman, Smith, Phuan, Tradtrantip, & Anderson, 2017).

No connexin-specific drugs targeting Cx43 have received approval yet, but it have been targeted in development of treatment for epidermal wounds using antisense

oligodeoxynucleotide (AsODN), which is blocking Cx43 expression (Laird and Lampe 2018).

Biologic medicine Eculizumab is under consideration for treatment of neuromyelitis optica.

There are data showing that patients who received eculizumab had significantly lower risk of relapse than placebo group (Pittock et al., 2019).

Understanding the reciprocal regulation of AQP4 and Cx43 could open new avenues for interventions in neuropathologies and guide development of pharmaceutical agents.

1.4.2 Aims of the thesis

This master project is based on indications of the interconnection and reciprocal regulation between AQP4 and astrocytic and gap junctions in the brain.

This study is a continuation of recent findings in the host laboratory showing that an increase in the number of astrocytic gap junctions in the hippocampus of mice with target deletion of AQP4 (Katoozi et al. 2017).

(25)

14 The aim was to investigate whether it is the localization of AQP4 in astrocyte endfeet or its expression level which is responsible for the observed increase in the number of astrocytic gap junctions in AQP4 KO. We therefore studied two models:

1) Mice with heterozygous and homozygous deletion of AQP4 where AQP4 expression is halved (Het) or completely lost (KO) at protein and mRNA level

2) Mice with genetic deletion of syntrophins - β1-syntrophin knockout and αβ-syntrophin knockout (KO) mice, where AQP4 is mislocalized from the perivascular membranes, without affecting its protein and mRNA expression levels.

(26)

15

2 Methods

2.1 Animals

Samples from adult (3- to 6-month old) mice were used in this study with:

1. Targeted deletion of the genes encoding β1-syntrophin or both α and β-syntrophin.

The αβ-syntrophin double knockout (αβ-syn KO) mice were created by crossing

individual knockouts of α-syntrophin and β-syntrophin (Kim et al. 2018; Rao et al. 2019;

Adams et al. 2000).

2. Homozygous AQP4 knockout (AQP4 KO) and AQP4 heterozygous knockout (Het).

Wild type mice were used as controls. Different forms of a gene are referred to as alleles.

Diploid organisms carry two copies of each gene, they can have identical alleles - be homozygous for a gene, or can carry different alleles - be heterozygous for a gene (Lodish H 2000). Homozygous deletion refers to the loss of both alleles (100% loss of gene expression), while heterozygous deletion refers to the single allele deletion typically resulting in halved expression of the gene (Liu et al. 2008). We used adult male mice of the C57/BL6J strain (Salameh 2019). The AQP4 Het and WTs were littermates, whereas AQP4 KO samples were from a different litter of previously generated AQP4 KO mice .which can have some impact on results.This will be discussed later in the discussion section.

2.2 Perfusion fixation

Mice were anesthetized and transcardially perfused with a solution of 4% formaldehyde (FA) in 4% FA and 0.1% glutaraldehyde in 0.1 M PB for immunogold cytochemistry. The brains were taken out and post-fixed overnight in the fixation solution and stored in a 1:10 dilution of the same solution in 0.1 M PB (Katoozi et al. 2017).

(27)

16

2.3 Electron microscopy

Electron microscopy produces images of the ultrastructure of cellular components and have a much higher resolution than images taken with light microscopes. Properties of electron microscopes (e.g. high-vacuum) requires specific preparation and staining techniques to provide us images showing the ultrastructure of cells and tissues.

Combined with immunocytochemical methods, electron microscopy is a reliable and useful method to label single, specific protein molecules providing a connection of the ultrastructure and its function (Kaech 2002).

Figure 8 shows resemblance of an transmission electron microscope (TEM) with a light microscope. An electron beam is created with help of a heated filament. Electrons are further speed up with high voltages (60 – 1200 kV decided by the type of TEM) and guided through the microscope column using electromagnetic lenses. The electron beam goes through and interacts with the sample giving us an image. The image can be monitored on a

phosphorescent screen or specially designed CCD camera and recorded on the computer connected with the microscope by a special software.

Figure 8 (Kaech 2002) – showing similarity between electron microscope (EM) and the conventional light microscope. Main difference is that EM use an electron beam instead of visible light.

(28)

17 Figure 9 – EM picture of brain blood vessel taken at 8.000x magnification;

E; endothelial cells, L; vessel lumen, arrow is pointing at dots representing gold labeling of gap junctions, P; pericyte

(29)

18 Figure 10 – EM picture of brain blood vessel taken at 20.500x magnification. The arrow is pointing at the immunogold particles marking the gap junction E; endothelial cells, L; vessel lume, P; pericyte

Brain tissue from mouse hippocampus was been labeled by the immunogold method and I peformed image acquisition in the electron microscope. Pictures were analyzed by

quantifying gap junctions and significance interpreted using statistic tests.

A TECNAI 12 TEM electron microscope was used to examine the sections at 80 kV. 20-30 capillaries depending on the sample quality were randomly selected at low magnification (8 000x) and at 20 500x magnification.

(30)

19 The size of the blood vessels was used as the elimination criterion. The aim was to take

images of the capillaries that are entirely in the field of view at a magnification of 8 000 x.

2.4 Immunogold method

Immunogold method is combining the power of the electron microscope with the localizing accuracy of immunocytochemistry. By using gold particles as markers in this method, proteins can be identified at notable precision followed by opportunities for its quantitation (Amiry-Moghaddam and Ottersen 2013).

Figure 11 – showing principles behind a two-step, postembedding immunogold method.

Triangles appearing on the surface of ultrathin section represents the antigen which the primary antibody (A) is up against. The secondary antibody (B) is connected to a colloidal gold particle (Au). Effective antibody diameter is 8 nm (Amiry-Moghaddam and Ottersen 2013).

(31)

20 When labeled by the immunogold method, the individual Connexin 43 proteins are labeled with a gold particle. In theory, each gap junction hemichannel contains six Cx43 proteins and each gap junction twelve Cx43 proteins. In practice, not all these proteins are present in each section containing gap junctions. Thus, our inclusion criteria for GJ were the following: all apparent gap junctions at high magnification (20 500x) were counted, regardless of distance from the capillary, but only if the membranes were clear and a minimum of 4 gold particles are detected in the gap junction.

In this study, the hippocampus was the region of interest. This is based on findings that increase in connexin gap junction formation in AQP4 KO mice was more prominent in hippocampus than in parietal cortex (Katoozi et al. 2017). Previously have also been reported regional differences in the regulation of expression astrocytic gap junctions (Ball et al. 2011;

Aberg et al. 2000).

The researchers performing the analysis were always blinded to the genotype of the animals.

The quantitation was performed separately by two investigators, the numbers compared and a consensus reached according to the established inclusion criteria.

Brains were harvested and cut into 0.5–1 mm tissue blocks; hippocampus war dissected, cryo- protected, quick-frozen in liquid propane (−170 °C), and subjected to freeze substitution.

Specimens were embedded in methacrylate resin (Lowicryl HM20) and polymerized by UV light below 0 °C (Amiry-Moghaddam et al. 2005).

Sections were incubated overnight with primary antibody diluted in Tris-buffered saline with Triton X-100 (TBS-T) with 2% (w/v) human serum albumin (HSA) at RT.

Sections were then rinsed with TBS-T (3 × 10 min) and incubated in TBS-T containing 2%

(w/v) HSA (10 min) before incubation with goat anti-rabbit (GAR) secondary antibody coupled to 15 nm colloidal gold particles for 2 h. The sections were then rinsed with distilled H2O, dried and contrasted with 2% uranyl acetate for 90 s and 0.3% lead citrate for 90 s (Katoozi et al. 2020).

The sections were examined with a TECNAI 12 transmission electron microscope. In each animal, 20-30 images of capillaries from hippocampus were randomly taken at the

magnification of 8.000× and 20.500×. The investigators were blinded to the genotype of the animals when taking and analyzing the pictures.

(32)

21

2.5 Statistical analysis

The SPSS software was used for statistical analysis (version 26, IBM). The purpose of the statistical analysis was to analyze whether there were significant differences in the number of gap junctions between the different groups. ANOVA with Fisher’s Least Significant

Difference (LSD) post hoc test was performed to determine whether the probability (P)-value is 0.05 or less, which we considered to be the significance level.

(33)

22

3 Results

3.1 Non-significant difference in GJ number

between AQP4 he and AQP4 KO mice compared to

WT

(34)

23 Figure 12 - a), c), e) shows pictures taken at 8.000x magnification respectively from from WT, AQP4 Het and AQP4 KO samples while b), d), f) shows pictures taken at 20.500x

magnification respectively from from WT, AQP4 Het and AQP4 KO samples. Arrows are pointing at gold labeling of the gap junctions. g) shows statistical analysis; E; endothelial cells, L; vessel lumen

(35)

24 Figure 13 – Table containing information number of gap junctions in individual blood vessels were used for further processing and analysis of data.

Statistical analysis showed no significant differences between the number of gap junctions in the three groups in this experiment. This was surprising as Katoozi et al reported an increase in the number of gap junctions in AQP4 KO compared to WT (Katoozi et al. 2017). Possible explanations will be discussed under discussion section.

(36)

25

3.2 Increased number of GJ in αβ -syntrophin

KO mice compared to β1 KO and WT

(37)

26 Figure 14 - a), c), e) shows pictures taken at 8.000x magnification from WT, β1 KO and αβ KO samples while b), d), f) shows pictures taken at 20.500x magnification from WT, β1 KO and αβ KO samples. Arrows are pointing at gold labeling of the gap junctions. g) shows statistical analysis; E; endothelial cells, L; vessel lumen; * - significant increase of

expression of gap junctions in αβ-syntrophin KO mice compared to both β1-syntrophin KO model and the wild type.

(38)

27 Figure 15 – Table containing information number of gap junctions in individual blood vessels were used for further processing and analysis of data.

The highest number of gap junctions was observed in αβ-syntrophin knockout (KO) mice.

Statistical tests showed significant increase of expression of gap junctions in αβ-syntrophin knockout mice compared to both β1-syntrophin knockout model and the wild type.

On the other hand, increased expression of gap junctions in β1-syntrophin KO compared to the wild type is not shown to be significant.

(39)

28

4 Discussion

4.1 Methodological considerations

The main purpose of this study was to investigate the effect of either AQP4 knockout or knockout of AQP4-anchoring molecules (syntrophins) on number of astrocytic gap junctions.

The reciprocity between AQP4 and gap junctions has been explored in two separate

experiments, only one of which has been able to produce sufficiently clear results as expected.

Genetic deletion of αβ-syntrophin, the molecules responsible for anchoring AQP4, led to a significant increase in the number of gap junctions. On the other hand, deletion of AQP4 did not produce the expected results, ie. did not lead to the expected significant increase in the number of gap junctions – an doubling of gap junctions in hippocampus of AQP4 KO mice compared to wild type has been previously shown (Katoozi et al. 2017). However, the validity and accuracy of the results obtained in AQP4 project for several reasons can be brought into question. In short, it is likely that the impact of animals from different litters was decisive for the inconsistency of these results. As previously stated, mice used in this part of project were not littermates. The AQP4 Het and WTs were littermates, but the AQP4 KO samples used in this study were from a different litter of animals in an earlier breeding setup. There are some data showing that relatively small differences in breeding conditions could have significant effects on certain aspects of brain development in mice (Wahlsten 1982). This part of the project leaves few questions open and leaves room for future analysis. In future studies, it would be advisable to examine the effect of AQP4 deletion on animals from the same litter.

The biggest challenge in the methodological part was to reach consensus on counting gap junctions. There is no universal rule that accurately defines the distance of gap junctions from a blood vessel. Non-specific staining also created great interference. Consequently, it was necessary to rely, for the most part, on the anatomical structure.

As a precautionary measure, the two researchers counted independently of each other. The researchers performing the analysis were blinded to the genotype of the animals. However, to avoid individual differences in the interpretation of distance from the blood vessel, the

(40)

29 following rule was adopted: all GJ at high magnification (zoom) were counted, regardless of distance from the capillary, but only if the membrane is clear and there are 4 or more dots.

Figure 16 – Boxplot shows interpretation of results by different researchers in AQP4 Het/knockout model. The different colors represent respectively results from researcher 1, researcher 2 and the final results about which agreement was made. On the x axis we can see the grid number and on the y axis the average number of GJ per micrograph. A1-A4 are WT, B1-B4 are AQP4 Het and C1-C4 are AQP4 KO samples

There were more challenges in performing this analysis. The idea was to take about 20-30 pictures of capillaries from each sample. Four samples were used for three different genotypes for each project. In practice, however, in some cases it was difficult to find enough capillaries, because simply some of the samples were not stable enough during microscopy. However, duplicates of each sample were made as a precaution. This turned out to be very useful, because duplicates were used instead of two samples in the AQP4 project.

(41)

30 Figure 17 – Boxplot shows interpretation of results by different researchers in syntrophin knockout model. The different colors represent respectively results from researcher 1,

researcher 2 and the final results about which agreement was made. On the x axis we can see the grid number and on the y axis the average number of GJ per micrograph. A1-A4 are WT, B1-B4 are β1 KO and C1-C4 are αβKO samples.

(42)

31

4.2 AQP4 heterozygous/knockout model

This experiment showed no significant difference between AQP4 heterozygous and knockout model versus wild type. These results should be interpreted with caution.

One of the central questions in this study is why no significant increase in the number of gap junctions was found in AQP4 knockout models. Earlier it was found that targeted deletion of Aqp4 promotes the formation of gap junctions in astrocytes (Katoozi et al. 2017).

Accordingly, the number of gap junctions in the AQP4 knockout model is expected to increase.

The expected increase did not take place, and a likely explanation is because the AQP4 KO samples we used in the experiment were not from the same litter as WT and AQP4 Het (not

"litter mates"). This can be considered as a problem in basic research (animals from different labs could have greater differences in protein expression than expected). This could be

influenced by differences in breeding conditions which may have significant effects on certain aspects of brain development in mice (Wahlsten 1982).

Certainly, in future analyzes this element should be considered ahead of time.

In short, this part of the analysis failed to answer the questions that were raised as the goal of this thesis.

However, we did not find a significant difference in the number of gap junctions between the WT and Het mice, which were littermates. Unpublished data from our laboratory have shown that the expression level of AQP4 is significantly reduced (to about 40%) in the Het brains compared to the WT. The fact that we didn’t see a significant increase in the number of gap junctions in Het brains might indicate that the increase in the number of gam junctions observed by Katoozi et al in the AQP4KO mice might not be due to the changes in the AQP4 expression level, but rather its loss from the perivascular membrane domains. This need to be explored in the future studies.

(43)

32

4.3 Syntrophin knockout model

This part of the analysis was based on association between perivascular AQP4 expression and α-syntrophin that has been previously demonstrated (Neely et al., 2001). Due to the role of syntrophin in anchoring AQP4 (Amiry-Moghaddam, Frydenlund, and Ottersen 2004) and the hypothesis of the mutual regulation of AQP4 levels and gap junctions (Katoozi et al. 2017;

Katoozi et al. 2020), one might expect an increase in the number of gap junctions in the syntrophin knockout models.

As stated above, the challenge was to determine the criteria for quantification, as well as the different quality of the samples used in microscopy. The animals used in the syntrophin KO model experiments were all littermates or closely related mice, including WT animals.

In this project, we found that there was not a significant increase in the number of gap junctions in β1-syntrophin knockout model compared to the wild type. It has been indicated that β1-syntrophin might also play a role in the anchoring of aquaporin 4 (Puwarawuttipanit et al. 2006), which in turn would theoretically lead us to expect an increased number of gap junctions. However, the results of this analysis do not support this hypothesis. Also, one part of the premise about the possible role of β1-synthrophin may be based on the findings of a study of its role in the retina. Deletion of this syntrophin caused a loss of Kir4.1 potassium channels from Müller cell endfeet in the retina (Rao et al. 2019). However, it is possible that β1-synthrophin may not play an as central role in brain AQP4-anchoring compared to retinal AQP4 anchoring..Recent data by Katoozi, Rao et al have shown that deletion of β1-

syntrophin causes a partial loss of AQP4 from perivascular retinal Müller cell endfeet. It has been suggested that upregulation of perivascular α1-syntrophin compensates effect of β1- syntrophin deletion. (Katoozi, Rao, et al., 2020). This kind of regulation could be explaining no influence of β1-syntrophin deletion on the number of GJ.

The comparison of the groups showed significant increase of expression of gap junctions in the αβ-syntrophin knockout mice compared to both β1-syntrophin knockout model and the wild type. The significance of these results was confirmed by statistical analysis.

When both α and β are removed, compensation mechanisms or reciprocal upregulation of either α and β1-syntrophin is not possible. This could be why virtually no AQP4 will be anchored perivascularly, and therefore many more gap junctions are formed in the membrane.

(44)

33 Analysis of αβ-syntrophin KO mice have shown that the perivascular AQP4 is completely lost from perivascular endfeet without affecting its total expression levels according to

unpublished data from host laboratory.

In summary, the results that showed an increased number of gap junctions in αβ-syntrophin knockout models were consistent with the expected results.

4.4 General discussion of the findings

The possible reciprocal regulation between gap junctions and AQP4 is of great importance for understanding the connection between the two and for better understanding their function and their role in treating of several diseases.

Interconnection between AQP4 and Connexin 43/gap junctions might actually be more complicated than expected basic hypothesis of this paper. Recent work published by research group from host laboratory has shown a much more complex relationship between the two. In this study, deletion of astrocytic connexins Cx43 and Cx30 is shown to leads to a reduction of perivascular AQP4, but isoform expression analysis shows that the level of the predominant AQP4 M23 isoform is reduced, while the levels of M1, and the alternative translation AQP4ex isoform protein levels are increased (Katoozi et al., 2020). Future investigations should further examine the roles of isoforms and their potentials as targets for specific drugs (Li et al. 2019; Mylvaganam et al. 2014).

Both AQP4 and Cx43 play important roles in CNS homeostasis as well as in the development of various diseases and disorders. In particular AQP4 is a significant potential target in the study of possible drugs that could be used in the treatment of brain diseases.

The main challenges in the development of AQP4 drugs are that they show many homologous AQP isoforms, wide distribution in tissues and function, possible side effects in the CNS and other tissues, and the need for high blood-brain barrier permeation. Despite considerable effort, small molecules affecting AQP4 have not been developed (Verkman et al. 2017).

(45)

34 However, biologic medicine Eculizumab is under consideration for treatment of neuromyelitis optica, an autoimmune inflammatory disease of the CNS initiated by binding of anti-AQP4 autoantibodies to astrocytic AQP4 (Pittock et al. 2019; Verkman et al. 2017).

Connexin 43 could also in the future become significant target in development of epilepsy medicines (Li et al. 2019; Mylvaganam et al. 2014). Currently there are treatment targeting Connexin 43 which is under development for epidermal wounds using antisense

oligodeoxynucleotide (AsODN). This treatment is based on blocking of Cx43 expression (Laird and Lampe 2018).

Regardless of the potential difficulties and risks in the development of therapies that would have AQP4 and Cx43 as a target, it should be recognized the great potential. However, more research is needed on these molecules, which are most likely in a reciprocal relationship.

(46)

35

5 Conclusion

The purpose of this study was to determine the association between AQP4 water channels in the brain and their anchoring proteins syntrophins with the number of astrocytic gap

junctions. Two aspects have been studied, AQP4 deletion and syntrophin deletion.

We have shown significant increase of expression of gap junctions in αβ-syntrophin knockout (KO) mice compared to the wild type. This confirms hypothesis that deletion of syntrophin and subsequent loss of AQP4 in perivascular membrane domains leads to an increased number of gap junctions in the same domains.

The other part of the study leaves much more uncertainty and material for future research. In any case, the interrelationships between AQP4, syntrophins, and gap junctions could be even more complicated than expected.

Importance of the molecules studied in this study as a target in the treatment of various diseases and disorders in the CNS is under development and is expected to be confirmed in the years to come.

(47)

36

6 Appendix

6.1 List of abbreviations

ANOVA: Analysis of variance AQP: Aquaporin

AQP4: Aquaporin-4

AsODN: Antisense oligodeoxynucleotide BBB: Blood-brain barrier

CCD: Charge-coupled device CNS: Central nervous system Cx: Connexin

dKO: Double knockout EM: Electron microscope FA: Formaldehyde GJ: Gap junctions Het: Heterozygous KO: Knockout

LSD: Fisher’s Least Significant Difference (LSD) NMO: Neuromyelitis Optica

NPA: Aspargine‐proline‐alanine

(48)

37 ODDD: Oculodentodigital dysplasia

PB: Phosphate buffer

SPSS: Statistical Package for the Social Sciences TEM: Transmission electron microscope

WT: Wild type

(49)

38

6.2 TEM procedyre

The correct way to extract the specimen holder from the Compustage

Since the compustage is made of high precision mechanics, inserting and retracting the specimen is somewhat delicate operation. To ensure trouble free operation, please follow the instructions to avoid damage to the airlock, the specimen holder or the specimen stage.

Please note the following general principles:

1. Read the full handling instructions before you start.

2. Never use excessive force, it’s a sign that something is wrong.

3. Never extract the holder while the airlock is being pumped.

4. Always carry out the full procedure, never leave the specimen holder in the retracted position.

Removing the holder

1. Close the column valves.

2. Make sure the specimen holder is in a safe position, reset all axes to zero, go to the Stage folder and in the control folder press the Holder button. The red compustage light should be off (do not remove the holder if the red light is on).

3. Carefully extract the holder form the compustage. Press on the purple plate with your fingers during the whole extraction procedure.

Press the Holder button to set all axes to zero. Do this EVERY time the holder is removed from the Compustage.

(50)

39

References

Aberg, N. D., B. Carlsson, L. Rosengren, J. Oscarsson, O. G. Isaksson, L. Ronnback, and P.

S. Eriksson. 2000. 'Growth hormone increases connexin-43 expression in the cerebral cortex and hypothalamus', Endocrinology, 141: 3879-86.

Ackerman, S. 1992. 'Discovering the Brain.' in (National Academies Press (US): Washington (DC)).

Adams, M. E., M. H. Butler, T. M. Dwyer, M. F. Peters, A. A. Murnane, and S. C. Froehner.

1993. 'Two forms of mouse syntrophin, a 58 kd dystrophin-associated protein, differ in primary structure and tissue distribution', Neuron, 11: 531-40.

Adams, Marvin E., Neal Kramarcy, Stuart P. Krall, Susana G. Rossi, Richard L. Rotundo, Robert Sealock, and Stanley C. Froehner. 2000. 'Absence of α-Syntrophin Leads to Structurally Aberrant Neuromuscular Synapses Deficient in Utrophin', Journal of Cell Biology, 150: 1385-98.

Adeeb, Nimer, Martin M. Mortazavi, Aman Deep, Christoph J. Griessenauer, Koichi

Watanabe, Mohammadali M. Shoja, Marios Loukas, and R. Shane Tubbs. 2013. 'The pia mater: a comprehensive review of literature', Child's Nervous System, 29: 1803-10.

Agre, P. 2004. 'Aquaporin water channels (Nobel Lecture)', Angew Chem Int Ed Engl, 43:

4278-90.

Ahn, A. H., C. A. Freener, E. Gussoni, M. Yoshida, E. Ozawa, and L. M. Kunkel. 1996. 'The three human syntrophin genes are expressed in diverse tissues, have distinct

chromosomal locations, and each bind to dystrophin and its relatives', J Biol Chem, 271: 2724-30.

Ahn, Andrew H., and Louis M. Kunkel. 1993. 'The structural and functional diversity of dystrophin', Nature Genetics, 3: 283-91.

Amiry-Moghaddam, M., D. S. Frydenlund, and O. P. Ottersen. 2004. 'Anchoring of

aquaporin-4 in brain: molecular mechanisms and implications for the physiology and pathophysiology of water transport', Neuroscience, 129: 999-1010.

Amiry-Moghaddam, M., T. Otsuka, P. D. Hurn, R. J. Traystman, F. M. Haug, S. C. Froehner, M. E. Adams, J. D. Neely, P. Agre, O. P. Ottersen, and A. Bhardwaj. 2003. 'An alpha- syntrophin-dependent pool of AQP4 in astroglial end-feet confers bidirectional water flow between blood and brain', Proc Natl Acad Sci U S A, 100: 2106-11.

Amiry-Moghaddam, M., and O. P. Ottersen. 2003. 'The molecular basis of water transport in the brain', Nat Rev Neurosci, 4: 991-1001.

Amiry-Moghaddam, M., A. Williamson, M. Palomba, T. Eid, N. C. de Lanerolle, E. A.

Nagelhus, M. E. Adams, S. C. Froehner, P. Agre, and O. P. Ottersen. 2003. 'Delayed K+ clearance associated with aquaporin-4 mislocalization: phenotypic defects in brains of alpha-syntrophin-null mice', Proc Natl Acad Sci U S A, 100: 13615-20.

Amiry-Moghaddam, Mahmood, Heidi Lindland, Sergey Zelenin, Bjørg Å. Roberg, Brigitta B.

Gundersen, Petur Petersen, Eric Rinvik, Ingeborg A. Torgner, and Ole P. Ottersen.

2005. 'Brain mitochondria contain aquaporin water channels: evidence for the expression of a short AQP9 isoform in the inner mitochondrial membrane', The FASEB Journal, 19: 1459-67.

Amiry-Moghaddam, Mahmood, and Ole Petter Ottersen. 2013. 'Immunogold cytochemistry in neuroscience', Nature Neuroscience, 16: 798.

(51)

40 Anand, Kuljeet Singh, and Vikas Dhikav. 2012. 'Hippocampus in health and disease: An

overview', Annals of Indian Academy of Neurology, 15: 239-46.

Attwell, D., A. Mishra, C. N. Hall, F. M. O'Farrell, and T. Dalkara. 2016. 'What is a pericyte?', J Cereb Blood Flow Metab, 36: 451-5.

Ball, K. K., L. Harik, G. K. Gandhi, N. F. Cruz, and G. A. Dienel. 2011. 'Reduced gap

junctional communication among astrocytes in experimental diabetes: contributions of altered connexin protein levels and oxidative-nitrosative modifications', J Neurosci Res, 89: 2052-67.

Ballabh, P., A. Braun, and M. Nedergaard. 2004. 'The blood-brain barrier: an overview:

structure, regulation, and clinical implications', Neurobiol Dis, 16: 1-13.

Bhat, H. F., M. E. Adams, and F. A. Khanday. 2013. 'Syntrophin proteins as Santa Claus:

role(s) in cell signal transduction', Cell Mol Life Sci, 70: 2533-54.

Binder, D. K., E. A. Nagelhus, and O. P. Ottersen. 2012. 'Aquaporin-4 and epilepsy', Glia, 60:

1203-14.

Boison, Detlev, and Christian Steinhäuser. 2018. 'Epilepsy and astrocyte energy metabolism', Glia, 66: 1235-43.

Brown, D. 2017. 'The Discovery of Water Channels (Aquaporins)', Ann Nutr Metab, 70 Suppl 1: 37-42.

Castro Dias, Mariana, Caroline Coisne, Pascale Baden, Gaby Enzmann, Lillian Garrett, Lore Becker, Sabine M. Hölter, Antonio Aguilar-Pimentel, Thure Adler, Dirk H. Busch, Nadine Spielmann, Kristin Moreth, Wolfgang Hans, Oana Amarie, Jochen Graw, Jan Rozman, Ildiko Radc, Frauke Neff, Julia Calzada-Wack, Birgit Rathkolb, Eckhard Wolf, Thomas Klopstock, Wolfgang Wurst, Johannes Beckers, Manuela Östereicher, Gregor Miller, Holger Maier, Claudia Stoeger, Stefanie Leuchtenberger, Valérie Gailus-Durner, Helmut Fuchs, Martin Hrabě de Angelis, Urban Deutsch, Britta Engelhardt, and Consortium German Mouse Clinic. 2019. 'Claudin-12 is not required for blood–brain barrier tight junction function', Fluids and Barriers of the CNS, 16:

30.

Clark, Robert E., and Larry R. Squire. 2013. 'Similarity in form and function of the hippocampus in rodents, monkeys, and humans', Proceedings of the National Academy of Sciences of the United States of America, 110 Suppl 2: 10365-70.

Contreras, J. E., H. A. Sanchez, L. P. Veliz, F. F. Bukauskas, M. V. Bennett, and J. C. Saez.

2004. 'Role of connexin-based gap junction channels and hemichannels in ischemia- induced cell death in nervous tissue', Brain Res Brain Res Rev, 47: 290-303.

Daneman, R. 2012. 'The blood-brain barrier in health and disease', Ann Neurol, 72: 648-72.

Ezan, P., P. Andre, S. Cisternino, B. Saubamea, A. C. Boulay, S. Doutremer, M. A. Thomas, N. Quenech'du, C. Giaume, and M. Cohen-Salmon. 2012. 'Deletion of astroglial connexins weakens the blood-brain barrier', J Cereb Blood Flow Metab, 32: 1457-67.

Frederic, Collignon, M. Wetjen Nicholas, A. Cohen-Gadol Aaron, D. Cascino Gregory, Parisi Joseph, B. Meyer Fredric, W. Richard Marsh, Roche Patrick, and D. Weigand

Stephen. 2006. 'Altered expression of connexin subtypes in mesial temporal lobe epilepsy in humans', Journal of Neurosurgery JNS, 105: 77-87.

Garbelli, R., C. Frassoni, D. F. Condorelli, A. Trovato Salinaro, N. Musso, V. Medici, L.

Tassi, M. Bentivoglio, and R. Spreafico. 2011. 'Expression of connexin 43 in the human epileptic and drug-resistant cerebral cortex', Neurology, 76: 895.

Gentile, Maria, and Luca Colucci D'Amato. 2018. 'Introductory Chapter: The Importance of Astrocytes in the Research of CNS Diseases.' in.

Halliday, G. 2017. 'Pathology and hippocampal atrophy in Alzheimer's disease', Lancet Neurol, 16: 862-64.

(52)

41 Hinson, S. R., S. J. Pittock, C. F. Lucchinetti, S. F. Roemer, J. P. Fryer, T. J. Kryzer, and V.

A. Lennon. 2007. 'Pathogenic potential of IgG binding to water channel extracellular domain in neuromyelitis optica', Neurology, 69: 2221-31.

Hinson, S. R., M. F. Romero, B. F. Popescu, C. F. Lucchinetti, J. P. Fryer, H. Wolburg, P.

Fallier-Becker, S. Noell, and V. A. Lennon. 2012. 'Molecular outcomes of

neuromyelitis optica (NMO)-IgG binding to aquaporin-4 in astrocytes', Proc Natl Acad Sci U S A, 109: 1245-50.

Iacovetta, C., E. Rudloff, and R. Kirby. 2012. 'The role of aquaporin 4 in the brain', Vet Clin Pathol, 41: 32-44.

Iadecola, C., and M. Nedergaard. 2007. 'Glial regulation of the cerebral microvasculature', Nat Neurosci, 10: 1369-76.

Jorgensen, N. R., S. T. Geist, R. Civitelli, and T. H. Steinberg. 1997. 'ATP- and gap junction- dependent intercellular calcium signaling in osteoblastic cells', The Journal of cell biology, 139: 497-506.

Kaech, Andres. 2002. 'An introduction to electron microscopy instrumentation, imaging and preparation', Cent. Microsc. Image Anal: 1-26.

Katoozi, S., N. Skauli, S. Rahmani, L. M. A. Camassa, H. B. Boldt, O. P. Ottersen, and M.

Amiry-Moghaddam. 2017. 'Targeted deletion of Aqp4 promotes the formation of astrocytic gap junctions', Brain Struct Funct, 222: 3959-72.

Katoozi, Shirin, Nadia Skauli, Soulmaz Zahl, Tushar Deshpande, Pascal Ezan, Claudia Palazzo, Christian Steinhäuser, Antonio Frigeri, Martine Cohen-Salmon, Ole Petter Ottersen, and Mahmood Amiry-Moghaddam. 2020. 'Uncoupling of the Astrocyte Syncytium Differentially Affects AQP4 Isoforms', Cells, 9: 382.

Kim, Min Jeong, Nicholas P Whitehead, Kenneth L Bible, Marvin E Adams, and Stanley C Froehner. 2018. 'Mice lacking α-, β1- and β2-syntrophins exhibit diminished function and reduced dystrophin expression in both cardiac and skeletal muscle', Human Molecular Genetics, 28: 386-95.

Kim, Yonghee, Jinhong Park, and Yoon Kyung Choi. 2019. 'The Role of Astrocytes in the Central Nervous System Focused on BK Channel and Heme Oxygenase Metabolites:

A Review', Antioxidants (Basel, Switzerland), 8: 121.

Kimelberg, Harold K., and Maiken Nedergaard. 2010. 'Functions of astrocytes and their potential as therapeutic targets', Neurotherapeutics : the journal of the American Society for Experimental NeuroTherapeutics, 7: 338-53.

Klatzo, Igor. 1967. 'Neuropathological Aspects of Brain Edema*', Journal of Neuropathology

& Experimental Neurology, 26: 1-14.

Laird, D. W., and P. D. Lampe. 2018. 'Therapeutic strategies targeting connexins', Nat Rev Drug Discov, 17: 905-21.

Lan, Y. L., J. J. Chen, G. Hu, J. Xu, M. Xiao, and S. Li. 2017. 'Aquaporin 4 in Astrocytes is a Target for Therapy in Alzheimer's Disease', Curr Pharm Des, 23: 4948-57.

Li, Q., Q. Q. Li, J. N. Jia, Z. Q. Liu, H. H. Zhou, and X. Y. Mao. 2019. 'Targeting gap

junction in epilepsy: Perspectives and challenges', Biomed Pharmacother, 109: 57-65.

Liu, Wennuan, Chunmei Carol Xie, Yi Zhu, Tao Li, Jishan Sun, Yu Cheng, Charles M.

Ewing, Sue Dalrymple, Aubrey R. Turner, Jielin Sun, John T. Isaacs, Bao-Li Chang, Siqun Lilly Zheng, William B. Isaacs, and Jianfeng Xu. 2008. 'Homozygous deletions and recurrent amplifications implicate new genes involved in prostate cancer',

Neoplasia (New York, N.Y.), 10: 897-907.

Lodish H, Berk A, Zipursky SL, et al. . 2000. 'Molecular Cell Biology. 4th edition. .' in.

Manley, Geoffrey T., Miki Fujimura, Tonghui Ma, Nobuo Noshita, Ferda Filiz, Andrew W.

Bollen, Pak Chan, and A. S. Verkman. 2000. 'Aquaporin-4 deletion in mice reduces

(53)

42 brain edema after acute water intoxication and ischemic stroke', Nature Medicine, 6:

159-63.

Mathiisen, T. M., K. P. Lehre, N. C. Danbolt, and O. P. Ottersen. 2010. 'The perivascular astroglial sheath provides a complete covering of the brain microvessels: an electron microscopic 3D reconstruction', Glia, 58: 1094-103.

Mylvaganam, Shanthini, Meera Ramani, Michal Krawczyk, and Peter L. Carlen. 2014. 'Roles of gap junctions, connexins, and pannexins in epilepsy', Frontiers in Physiology, 5.

Nag, Sukriti. 2011. 'Morphology and Properties of Astrocytes.' in Sukriti Nag (ed.), The Blood-Brain and Other Neural Barriers: Reviews and Protocols (Humana Press:

Totowa, NJ).

Nagy, J. I., D. Patel, P. A. Ochalski, and G. L. Stelmack. 1999. 'Connexin30 in rodent, cat and human brain: selective expression in gray matter astrocytes, co-localization with connexin43 at gap junctions and late developmental appearance', Neuroscience, 88:

447-68.

Nielsen, S., E. A. Nagelhus, M. Amiry-Moghaddam, C. Bourque, P. Agre, and O. P. Ottersen.

1997. 'Specialized membrane domains for water transport in glial cells: high- resolution immunogold cytochemistry of aquaporin-4 in rat brain', J Neurosci, 17:

171-80.

Pache, F., B. Wildemann, F. Paul, and S. Jarius. 2017. '[Neuromyelitis optica]', Fortschr Neurol Psychiatr, 85: 100-14.

Pannasch, U., L. Vargova, J. Reingruber, P. Ezan, D. Holcman, C. Giaume, E. Sykova, and N.

Rouach. 2011. 'Astroglial networks scale synaptic activity and plasticity', Proc Natl Acad Sci U S A, 108: 8467-72.

Parpura, Vladimir, and Alexei Verkhratsky. 2012. 'Astrocytes revisited: concise historic outlook on glutamate homeostasis and signaling', Croatian medical journal, 53: 518- 28.

Pisani, Francesco, Angelo Sparaneo, Carla Tortorella, Maddalena Ruggieri, Maria Trojano, Maria Grazia Mola, Grazia Paola Nicchia, Antonio Frigeri, and Maria Svelto. 2013.

'Aquaporin-4 autoantibodies in Neuromyelitis Optica: AQP4 isoform-dependent sensitivity and specificity', PloS one, 8: e79185-e85.

Pittock, S. J., A. Berthele, K. Fujihara, H. J. Kim, M. Levy, J. Palace, I. Nakashima, M. Terzi, N. Totolyan, S. Viswanathan, K. C. Wang, A. Pace, K. P. Fujita, R. Armstrong, and D.

M. Wingerchuk. 2019. 'Eculizumab in Aquaporin-4-Positive Neuromyelitis Optica Spectrum Disorder', N Engl J Med, 381: 614-25.

Preston, G. M., T. P. Carroll, W. B. Guggino, and P. Agre. 1992. 'Appearance of water channels in Xenopus oocytes expressing red cell CHIP28 protein', Science, 256: 385- 7.

Puwarawuttipanit, W., A. D. Bragg, D. S. Frydenlund, M. N. Mylonakou, E. A. Nagelhus, M.

F. Peters, N. Kotchabhakdi, M. E. Adams, S. C. Froehner, F. M. Haug, O. P. Ottersen, and M. Amiry-Moghaddam. 2006. 'Differential effect of α-syntrophin knockout on aquaporin-4 and Kir4.1 expression in retinal macroglial cells in mice', Neuroscience, 137: 165-75.

Rackauskas, M., V. Neverauskas, and V. A. Skeberdis. 2010. 'Diversity and properties of connexin gap junction channels', Medicina (Kaunas), 46: 1-12.

Rao, Shreyas B., Shirin Katoozi, Nadia Skauli, Stanley C. Froehner, Ole Petter Ottersen, Marvin E. Adams, and Mahmood Amiry-Moghaddam. 2019. 'Targeted deletion of β1- syntrophin causes a loss of Kir4.1 from Müller cell endfeet in mouse retina', Glia, 67:

1138-49.

(54)

43 Rash, J. E., T. Yasumura, C. S. Hudson, P. Agre, and S. Nielsen. 1998. 'Direct immunogold

labeling of aquaporin-4 in square arrays of astrocyte and ependymocyte plasma membranes in rat brain and spinal cord', Proc Natl Acad Sci U S A, 95: 11981-6.

Rouach, N., E. Avignone, W. Meme, A. Koulakoff, L. Venance, F. Blomstrand, and C.

Giaume. 2002. 'Gap junctions and connexin expression in the normal and pathological central nervous system', Biol Cell, 94: 457-75.

Saez, J. C., V. M. Berthoud, M. C. Branes, A. D. Martinez, and E. C. Beyer. 2003. 'Plasma membrane channels formed by connexins: their regulation and functions', Physiol Rev, 83: 1359-400.

Salameh, Maher Ayman. 2019. 'AQP4 Expression and Distribution in Mice With Heterozygous Targeted Deletion of Aqp4', Master thesis, University of Oslo.

Santello, Mirko, Nicolas Toni, and Andrea Volterra. 2019. 'Astrocyte function from

information processing to cognition and cognitive impairment', Nature Neuroscience, 22: 154-66.

Spray, David C., and Eliana Scemes. 2013. 'Gap Junction Proteins (Connexins, Pannexins, and Innexins).' in Gordon C. K. Roberts (ed.), Encyclopedia of Biophysics (Springer Berlin Heidelberg: Berlin, Heidelberg).

Sun, M. C., C. R. Honey, C. Berk, N. L. Wong, and J. K. Tsui. 2003. 'Regulation of aquaporin-4 in a traumatic brain injury model in rats', J Neurosurg, 98: 565-9.

Theis, Martin, and Christian Steinhäuser. 2013. 'Chapter 2 - Physiology and Function of Glial Gap Junctions in the Hippocampus.' in Ekrem Dere (ed.), Gap Junctions in the Brain (Academic Press: San Diego).

Unterberg, A. W., J. Stover, B. Kress, and K. L. Kiening. 2004. 'Edema and brain trauma', Neuroscience, 129: 1019-27.

van de Haar, Harm J., Saartje Burgmans, Paul A. M. Hofman, Frans R. J. Verhey, Jacobus F.

A. Jansen, and Walter H. Backes. 2015. 'Blood–brain barrier impairment in dementia:

Current and future in vivo assessments', Neuroscience & Biobehavioral Reviews, 49:

71-81.

Verkhratsky, Alexei, and Maiken Nedergaard. 2018. 'Physiology of Astroglia', Physiological reviews, 98: 239-389.

Verkman, A. S., A. J. Smith, P. W. Phuan, L. Tradtrantip, and M. O. Anderson. 2017. 'The aquaporin-4 water channel as a potential drug target in neurological disorders', Expert Opin Ther Targets, 21: 1161-70.

Wahlsten, D. 1982. 'Deficiency of corpus callosum varies with strain and supplier of the mice', Brain Res, 239: 329-47.

Walha, K., F. Ricolfi, Y. Bejot, M. Nonent, and D. Ben Salem. 2013. 'Hippocampus: a

"forgotten" border zone of brain ischemia', J Neuroimaging, 23: 98-101.

Wallraff, A., R. Kohling, U. Heinemann, M. Theis, K. Willecke, and C. Steinhauser. 2006.

'The impact of astrocytic gap junctional coupling on potassium buffering in the hippocampus', J Neurosci, 26: 5438-47.

Xu, Zhiqiang, Na Xiao, Yali Chen, Huang Huang, Charles Marshall, Junying Gao, Zhiyou Cai, Ting Wu, Gang Hu, and Ming Xiao. 2015. 'Deletion of aquaporin-4 in APP/PS1 mice exacerbates brain Aβ accumulation and memory deficits', Molecular

Neurodegeneration, 10: 58.

Yamamoto, T., A. Ochalski, E. L. Hertzberg, and J. I. Nagy. 1990. 'On the organization of astrocytic gap junctions in rat brain as suggested by LM and EM

immunohistochemistry of connexin43 expression', J Comp Neurol, 302: 853-83.

Yamamoto, T., J. Vukelic, E. L. Hertzberg, and J. I. Nagy. 1992. 'Differential anatomical and cellular patterns of connexin43 expression during postnatal development of rat brain', Brain Res Dev Brain Res, 66: 165-80.

(55)

44 Yamasaki, Ryo. 2018. 'Connexins in health and disease', Clinical and Experimental

Neuroimmunology, 9: 30-36.

Yeager, M., and B. J. Nicholson. 1996. 'Structure of gap junction intercellular channels', Curr Opin Struct Biol, 6: 183-92.

Yu, Qi-Jin, Hong Tao, Xin Wang, and Ming-Chang Li. 2015. 'Targeting brain microvascular endothelial cells: a therapeutic approach to neuroprotection against stroke', Neural regeneration research, 10: 1882-91.

Zador, Zsolt, Shirley Stiver, Vincent Wang, and Geoffrey T. Manley. 2009. 'Role of

Aquaporin-4 in Cerebral Edema and Stroke.' in Eric Beitz (ed.), Aquaporins (Springer Berlin Heidelberg: Berlin, Heidelberg).

Impact of the deletion of syntrophins and AQP4 on the prevalence of astrocytic gap junctions