Chloride Homeostasis in Central Neurons

Tushar Yelhekar

The Department of Integrative Medical Biology

Umeå 2016

Responsible publisher under swedish law: the Dean of the Medical Faculty

This work is protected by the Swedish Copyright Legislation (Act 1960:729)

ISBN: 978-91-7601-602-2

ISSN: 0346-6612

New Series No: 1861

Cover illustrations: Membrane current traces from preoptic neurons

Elektronisk version tillgänglig på http://umu.diva-portal.org/

Printed by: Print & Media, Umeå University

Umeå, Sweden 2016

i

Dedicated to my teachers

ii

iii

Table of Contents

List of papers v

Abbreviations vi

Abstract vii

Introduction 1

GABA- and glycine-receptor mediated neurotransmission 2

Receptor structure 2

Currents 3

Desensitization 4

Neurosteroids 4

Thermodynamics of Cl– movement 5

Cation chloride cotransporters (CCCs) 7

KCC2 7

NKCC1 9

Thermodynamics of KCC2 and NKCC1 10

Role of impermeant anions in setting [Cl–]i 11

Specific aims of the thesis 12

Methodological considerations 13

Cell and tissue preparation 13

Electrophysiological recording 14

Electrophysiological measuring of [Cl–]i 15

Cl– transporter activity and [Cl–]i homeostasis 16

Estimating relative Cl– leak by a novel method 16

Immunoblotting 17

Computational modeling 17

Results and Discussion 19

Project 1: Estimating [Cl–]i by electrophysiological methods 19

Project 2: Role of Cl– transporters and extracellular matrix in [Cl–]i

homeostasis 22

Project 3: Influence of Cl– leak conductance on [Cl–]i homeostasis 26

Project 4: Role of KCC2 under conditions with different

inhibitory or excitatory conductances 30

Overall summary 31

Acknowledgements 32

References 33

Paper I-IV

iv

v

List of papers

This doctorate dissertation is based on the following four papers which will be referred

to by their roman numbers:

I. Yelhekar TD, Druzin M, Karlsson U, Blomqvist E and Johansson S

(2016) How to Properly Measure a Current-Voltage Relation?—

Interpolation vs. Ramp Methods Applied to Studies of

GABAA Receptors. Frontiers in Cellular Neuroscience 10:10. doi:

10.3389/fncel.2016.00010

II. Yelhekar TD, Kuznetsova T, Malinina E, Ponimaskin E, Dityatev A,

Druzin M and Johansson S (2016) Extracellular Matrix Regulates

Neuronal Chloride Concentration via K+-Cl–-cotransporter 2.

Manuscript.

III. Yelhekar TD, Druzin M and Johansson S (2016) Contribution of Resting Conductance, GABAA-Receptor Mediated Miniature Synaptic

Currents and Neurosteroid to Chloride Homeostasis in Central

Neurons. Manuscript.

IV. Johansson S, Yelhekar TD and Druzin M (2016) Commentary:

Chloride Regulation: A Dynamic Equilibrium Crucial for Synaptic

Inhibition. Frontiers in Cellular Neuroscience 10:182. doi:

10.3389/fncel.2016.00182

vi

Abbreviations

CCC cation chloride cotransporter

ChABC chondroitinase ABC

[Cl–]i intracellular chloride concentration

[Cl–]o extracellular chloride concentration

CNS central nervous system

CSPG chondroitin sulphate proteoglycan

DFCl driving force for Cl– current

DFGABA driving force for current mediated by GABAARs

ECl Cl– equilibrium potential

ECM extracellular matrix

EGABA reversal potential of current mediated by GABAARs

eIPSC evoked inhibitory postsynaptic current

EK K+ equilibrium potential

F Faraday constant

GABA γ-aminobutyric acid

GABAAR γ -aminobutyric acid type A receptor

GABABR γ -aminobutyric acid type B receptor

GABACR γ -aminobutyric acid type C receptor

gCl Cl– conductance

GlyR glycine receptor

IGABA current mediated by GABAARs

ICl Cl– current

IPSC inhibitory postsynaptic current

I-V current-voltage

KCC2 K+-Cl–-cotransporter 2

mIPSC miniature inhibitory postsynaptic current

MPN medial preoptic nucleus

NKCC1 Na+-K+-Cl–-cotransporter 1

P permeability

R gas constant

sIPSC spontaneous inhibitory postsynaptic current

T temperature

Vm membrane potential

vii

Abstract

The overall aim of the present thesis is to clarify the control of intracellular chloride

homeostasis in central neurons, because of the critical role of chloride ions (Cl–) for

neuronal function. Normal function of the central nervous system (CNS) depends on

a delicate balance between neuronal excitation and inhibition. Inhibition is, in the

adult brain, most often mediated by the neurotransmitter γ-aminobutyric acid

(GABA). GABA may, however, in some cases cause excitation. GABA acts by

activating GABA type A receptors (GABAARs), which are ion channels largely

permeable to Cl–. The effect of GABAAR-mediated neuronal signaling - inhibitory or

excitatory - is therefore mainly determined by the Cl– gradient across the membrane.

This gradient varies with neuronal activity and may be altered in pathological

conditions. Thus, understanding Cl– regulation is important to comprehend neuronal

function. This thesis is an attempt to clarify several unknown aspects of neuronal Cl–

regulation. For such clarification, a sufficiently sensitive method for measuring the

intracellular Cl– concentration, [Cl–]i, is necessary. In the first study of this thesis, we

examined two electrophysiological methods commonly used to estimate [Cl–]i. Both

methods, here called the interpolation and the voltage-ramp method, depend on an

estimate of the Cl– equilibrium potential from the current-voltage relation of GABA-

or glycine-evoked Cl– currents. Both methods also provide an estimate of the

membrane Cl– conductance, gCl. With a combination of computational and

electrophysiological techniques, we showed that the most common (interpolation)

method failed to detect changes in [Cl–]i and gCl during prolonged GABA application,

whereas the voltage-ramp method accurately detected such changes. Our analysis also

provided an explanation as to why the two methods differ. In a second study, we

clarified the role of the extracellular matrix (ECM) for the distribution of Cl– across

the cell membrane of neurons from rat brain. It was recently proposed that immobile

charges located within the ECM, rather than as previously thought cation-chloride

transporter proteins, determine the low [Cl–]i which is critical to GABAAR-mediated

inhibition. By using electrophysiological techniques to measure [Cl–]i, we showed that

digestion of the ECM decreases the expression and function of the neuron-specific K+

Cl– cotransporter 2 (KCC2), which normally extrudes Cl– from the neuron, thus

causing an increase in resting [Cl–]i. As a result of ECM degradation, the action of

GABA may be transformed from inhibitory to excitatory. In a third study, we

developed a method for quantifying the largely unknown resting Cl– (leak)

conductance, gCl, and examined the role of gCl for the neuronal Cl– homeostasis. In

isolated preoptic neurons from rat, resting gCl was about 6 % of total resting

conductance, to a major part due to spontaneously open GABAARs and played an

important role for recovery after a high Cl– load. We also showed that spontaneous,

impulse-independent GABA release can significantly enhance recovery when the

GABA responses are potentiated by the neurosteroid allopregnanolone. In a final

commentary, we formulated the mathematical relation between Cl– conductance,

viii

KCC2-mediated Cl– extrusion capacity and steady-state [Cl–]i. In summary, the

present thesis (i) clarifies how well common electrophysiological methods describe

[Cl–]i and gCl, (ii) provides a novel method for quantifying gCl in cell membranes and

(iii) clarifies the roles of the ECM, ion channels and ion transporters in the control of

[Cl–]i homeostasis and GABAAR-mediated signaling in central neurons.

Keywords: chloride concentration, current-voltage relation, interpolation, voltage

ramp, reversal potential, [Cl–]i recovery, KCC2, extracellular matrix, GABAA

receptor, chloride leak conductance, neurosteroid

1

Introduction

Chloride (Cl–) is the most abundant physiological anion, involved in a number of

cellular functions, including pH regulation (Boron & Russell, 1983; Boron, 2004;

Ruffin et al., 2014), gene regulation (Succol et al., 2012; Valdivieso et al., 2016),

trans-epithelial ion transport (Hartzell, 2009; Sun et al., 2009) and cell-volume

regulation (Sardini et al., 2003; MacAulay et al., 2004; MacAulay & Zeuthen, 2010).

In addition, Cl– is crucial for neuronal signaling. Normal function of the central

nervous system (CNS) depends on a delicate balance between neuronal excitation and

inhibition, mediated by signals in the synaptic contacts between neurons. In the brain,

γ-aminobutyric acid (GABA) is by far the most common neurotransmitter that

mediates inhibition. GABA binds to and activates GABA type A receptors

(GABAARs) in the neuronal cell membrane. They are ligand-gated ion channels

largely permeable to Cl– and to a small extent to bicarbonate (HCO3–) (Bormann et

al., 1987). The strength and polarity of GABAAR-mediated signaling is mainly

determined by the number of open ion channels and by the Cl– gradient across the

membrane. In mature neurons, the intracellular Cl– concentration ([Cl–]i) is low, often

5 - 10 mM, due to Cl– extrusion mechanisms (Rivera et al., 1999). As a consequence,

activation of GABAARs leads to a net influx of Cl–, hyperpolarization of the

membrane and neuronal inhibition, that is, a reduced tendency to generate impulses.

Since the inhibitory signaling is associated with Cl– flux through the membrane, it

will also be associated with changes in [Cl–]i. Therefore, to maintain functional

signaling properties, neurons must be equipped with mechanisms to restore [Cl–]i.

There are, however, several conditions under which GABA may excite rather than

inhibit affected neurons. First, this may occur early in development, when [Cl–]i is

much higher than in mature neurons (Ben-Ari et al., 1989; Ben-Ari, 2002). In addition,

there are several pathological conditions where the GABAAR-mediated inhibition is

compromised and an associated maintained change of [Cl–]i (also called [Cl–]i shift)

is evident. These conditions include epilepsy (Kaila & Miles, 2010; Khirug et al.,

2010; Kaila et al., 2014b; Karlócai et al., 2015), chronic pain (Coull et al., 2003; De

Koninck, 2009; Gagnon et al., 2013), autism (Lemonnier & Ben-Ari, 2010; Tyzio et

al., 2014), stress (Hewitt et al., 2009; Maguire, 2014; Bains et al., 2015), and spinal

cord injury (Boulenguez et al., 2010). Despite many studies, the mechanisms and

factors involved in Cl– regulation are not well understood. Thus, for instance, the

causes of the [Cl–]i shifts during transitions to pathological states are not clear (Kaila

et al., 2014a; Wu et al., 2015). As a consequence, understanding Cl– regulation is not

only important to comprehend normal neuronal function but also to understand the

etiology of disease, necessary for the development of therapeutic drugs.

The overall aim of my Ph.D. research project was to clarify the intracellular chloride

homeostasis in central neurons. The present thesis is based on three main projects. In

2

the first project (paper I), we compared two common electrophysiological methods to

understand how [Cl–]i can best be measured and how well the methods reflect the Cl–

conductance, which is determined by the number of Cl–-permeable channels in the

neuronal cell membrane. In the second project (paper II), we studied the role of two

potentially critical factors for setting [Cl–]i and controlling synaptic inhibition: (i) Cl–

-transporting membrane proteins, previously thought to play an important role, but

recently questioned, and (ii) the extracellular matrix (ECM) that normally surrounds

the brain cells, previously not thought directly important for signaling, but recently

shown to influence [Cl–]i. Besides transporters, passive flux through ion channels is

expected to influence the resting [Cl–]i. Therefore, in the third project (paper III) we

developed a new method and measured the Cl– background, “leak”, conductance and

its influence on Cl– homeostasis. We also, in a “general commentary” (paper IV)

clarified the role of a Cl– transporter under different conditions by using a

mathematical model and computer simulations.

GABA- and glycine-receptor mediated neurotransmission

Inhibitory neurotransmission is important for proper neuronal coding and CNS

function. In the adult CNS, GABA and glycine are the most common inhibitory

neurotransmitters. GABA is predominantly present in the brain, whereas glycine is

common in the spinal cord and brainstem. Besides GABAA receptors, GABA acts on

several other types of receptors such as GABAB and GABAC receptors. GABAB

receptors (GABABRs) are metabotropic membrane receptors that mediate neuronal

signaling via G-protein activation. They often underlie slow synaptic signals and are

mainly present in presynaptic sites, where the main function is to reduce

neurotransmitter release (Takahashi et al., 1998). A controversial third type, GABA

C receptors (GABACRs) are ion channels like GABAARs. GABACRs are now usually

considered to be a subtype of GABAARs (Olsen & Sieghart, 2008). Similarly to

GABAARs, glycine receptors (GlyRs) are ligand-gated ion channels, permeable to

anions like Cl– and HCO3–, and mediate fast neurotransmission (Eccles et al., 1977;

Bormann et al., 1987; Kaila et al., 1993).

Receptor structure: GABA- and glycine-gated ion channels are functionally and

structurally similar. They belong to the Cys-loop receptor superfamily. Other

members of this family are cation-selective nicotinic acetylcholine receptors and

serotonin type 3 receptors. GABAARs are assembled in a heteropentameric structure

from a large family of subunits (α1–6, β1–3, γ1–3, δ, ε, θ, π and ρ1–3) encoded by 19

different genes. The major native isoforms are composed of two α, two β and a single

γ, δ or ε subunit (Olsen & Sieghart, 2009; Gunn et al., 2015). GABACRs are

homopentameric receptors composed of ρ subunits. Each subunit contains four

transmembrane (TM1-4) domains (Figure 1A). An extracellular domain includes

GABA- and receptor modulator (see below) binding sites (Olsen & Tobin, 1990). A

3

prominent intracellular domain between TM3 and TM4 is a target of both

serine/threonine and tyrosine protein kinases (Moss et al., 1995; McDonald et al.,

1998; Moss & Smart, 2001). The central ion channel pore is solely formed by the

TM2 domains of each subunit (Macdonald & Botzolakis, 2009) (Figure 1B). Synaptic

GABAARs contain a γ subunit and, compared with δ-subunit containing GABAARs,

display a lower affinity for GABA. Conversely, α4-, α5- and δ-subunit containing

receptors dominate in extrasynaptic sites (Farrant & Nusser, 2005).

Figure 1. Structure of Cys-loop receptor ion channels. A. Individual subunits contain four

transmembrane domains (TM1-4). B. A receptor is assembled in a pentameric structure and

the ion channel is formed by a ring of TM2 domains (blue). (Reproduced from Moss & Smart,

2001, with permission.)

Currents: Presynaptic release of GABA leads to activation of the postsynaptic

GABAARs and opening of the channels that mediate inhibitory postsynaptic currents

(IPSCs), also called phasic currents due to their transient nature. In common

terminology, spontaneous IPSCs (sIPSCs) refer to those generated in the absence of

external stimulation, miniature IPSCs (mIPSCs) refer to those generated

spontaneously when impulse generation is blocked, whereas evoked IPSCs (eIPSCs)

refer to those evoked by stimulation expected to cause presynaptic impulse generation

(Farrant & Nusser, 2005). Besides the action at the synaptic GABA receptors

responsible for the phasic currents, GABA may also escape from the synapse and

activate extrasynaptic GABAARs (Semyanov et al., 2003). Extrasynaptic receptors

have a higher affinity for GABA than their synaptic counterparts and they carry more

sustained, tonic currents (Brickley et al., 1996; Semyanov et al., 2004). Changes in

tonic current influence the neuronal action potential threshold and firing pattern

(Mitchell & Silver, 2003; Semyanov et al., 2004). As noted above, the currents

associated with GABAAR activation, whether phasic or tonic, are accompanied by

changes in Cl– concentrations that must be coped with to maintain normal function.

Although it is known that synaptic activation may be associated with a rise in [Cl–]i

4

(Thompson & Gähwiler, 1989), the influence of receptor activation on the neuronal

Cl– homeostasis in various conditions is still not well understood.

Desensitization: Although GABA-receptor channels are opened by GABA, they may

subsequently close during maintained exposure to GABA. This so-called

desensitization is a property shared with most other ligand-gated ion channels. It may

occur at different time scales and is thought crucial for synaptic communication. In

general, desensitization may be thought to limit the duration of synaptic influence, but

paradoxically the entrance to a desensitized state may imply prolonged synaptic

currents because of subsequent channel re-entry to the open state (Jones & Westbrook,

1995). The currents mediated by GABA- and glycine-gated channels, as commonly

studied under voltage-clamp conditions at steady membrane potentials, usually decay

with time at prolonged exposure to the agonist. This has been interpreted as reflecting

desensitization. However, recently Karlsson et al. (2011) showed that during

activation of GABAARs and GlyRs, the current decay is largely due to dramatic

changes in [Cl–]i whereas true desensitization is slow or even absent. True

desensitization, however, should be reflected by a decay of GABA/glycine-

mediated conductance. To clarify the properties of GABA- and glycine-gated

channels, critically important for synaptic inhibition, it is therefore crucial to be

able to estimate both [Cl–]i and channel-induced conductance. Because of different

conclusions in previous studies (cf e.g. Bianchi & Macdonald, 2002; Karlsson et al.,

2011), we, in paper I, clarified the consequences of the two most common

electrophysiological methods used to estimate conductance and ion concentrations.

As described below, the methods differ dramatically with respect to how correctly

these parameters are given.

Neurosteroids: A GABAAR protein complex has many binding sites, two for the

ligand GABA and several modulatory sites for benzodiazepines, barbiturates and

neurosteroids (Majewska et al., 1986; Barbaccia et al., 2001). Peripherally

synthesized neuroactive steroids may modulate neuronal activity by acting at

GABAARs (Strömberg et al., 2006; Herd et al., 2007). Some steroids are converted to

neuroactive forms by cells within the CNS. However, neurons and glial cells also

express enzymes for de novo synthesis of neurosteroids within the CNS (Corpéchot

et al., 1981; Mellon & Deschepper, 1993). Significant production of neuroactive

steroids or neurosteroids, with resulting changes in the steroid levels within the CNS,

is observed during specific developmental stages or physiological conditions, like

gestation, delivery (Concas et al., 1998; Luisi et al., 2000) and during stress (Purdy et

al., 1991), but also in pathological conditions, like epilepsy (Bäckström, 1976; Grosso

et al., 2003; Mareš et al., 2006). The locally synthesized neurosteroids act in a

paracrine manner by modulating GABAARs in neighboring CNS regions.

Endogenously produced neurosteroids including the progesterone metabolites 5α-

pregnan-3α-ol-20-one (3α,5α-THPROG or allopregnanolone) and 5β-pregnan-3α-ol-

5

20-one (3α,5β-THPROG), and the deoxycorticosterone metabolite 5α-pregnan-3α,21-

diol-20-one (3α,5α-THDOC) are positive modulators of GABAARs (Belelli &

Lambert, 2005; Lambert et al., 2009). Binding of a neurosteroid to a GABAAR leads

to changes in receptor conformation and channel kinetics (Majewska et al., 1986). The

effects depend on the neurosteroid concentration. For instance, at high or low

concentration, allopregnanolone directly activates or potentiates the action of GABA

at GABAARs, respectively (Haage & Johansson, 1999). Since allopregnanolone in

several ways may dramatically increase the GABAAR-mediated Cl– flux, it may

significantly affect the Cl– homeostasis. In the present work (paper III), the effect of

allopregnanolone on passive, channel-mediated recovery after a rise in [Cl–]i was

clarified.

Thermodynamics of Cl– movement

The current mediated by GABAARs (IGABA) can be described by Ohm’s law as a

product of conductance (gGABA; inverse of resistance) and the driving force (DFGABA),

that is, the difference between membrane potential (Vm) and reversal potential

(EGABA):

IGABA = gGABA (Vm – EGABA)

where, gGABA is the GABA-evoked conductance, which depends on the number of

open channels as well as the single-channel conductance. The reversal potential EGABA

is the potential at which no net GABA-mediated current occurs. GABAARs are

permeable to Cl– and HCO3– but the permeability to Cl– is 3 - 5 times higher than that

to HCO3–. Thus, EGABA is more influenced by Cl– than by HCO3

–, and in many

experimental conditions, the influence of HCO3– can be made negligible. In such

cases, EGABA will be equal to the equilibrium potential for Cl– (ECl) as described by

the Nernst equation:

ECl = RT/F ln([Cl–]i/[Cl–]o)

where, R is the gas constant, T is absolute temperature and F is Faraday’s constant. In

physiological conditions, however, EGABA is also influenced by HCO3–, so that

EGABA = (RT/F) ln((PCl[Cl–]i + PHCO3[HCO3–]i)/(PCl[Cl–]o + PHCO3[HCO3

–]o))

where P denotes the relative permeabilities for Cl– and HCO3–, as given by the

subscript. At high [Cl–]i, EGABA is then still relatively close to ECl, but at low [Cl–]i,

significant deviations in positive direction are caused by HCO3– (see e. g. Farrant &

Kaila, 2007).

6

As shown in Figure 2, the relation between ECl and [Cl–]i is logarithmic and suggests

that small changes in Cl– gradient may strongly affect ECl and therefore the driving

force (DFCl = Vm – ECl) that determines the direction of Cl– currents (ICl):

ICl = gCl (Vm – ECl) (1)

where gCl is Cl– conductance. Thus, the effect of ECl on GABAAR-mediated currents

and neuronal excitability is dramatic:

1) When ECl is more negative than Vm, activation of GABAARs will cause

influx of Cl−, membrane hyperpolarization and thereby neuronal inhibition.

GABAAR-mediated inhibition is commonly observed in the mature brain. To maintain

a low ECl, mechanisms to keep [Cl–]i low are required.

Figure 2. Relationship between intracellular chloride concentration [Cl–]i and equilibrium

potential for Cl– (ECl). Small changes in [Cl–]i may lead to profound changes in ECl and in the

effect of GABA action.

2) When ECl is more positive than Vm, GABAAR activation will lead to

outward flux of Cl− ions and depolarization of the membrane. GABA-mediated

depolarization does not always cause excitation. Subthreshold depolarization may

impose a strong inhibitory action, as observed in sensory afferent neurons (Khazipov

et al., 2014; Hammond, 2015). If ECl is more positive than Vm and above threshold for

action potential generation, GABAAR activation will likely, at least temporarily, be

excitatory and thus increase the firing probability. To constantly maintain a high ECl,

as commonly observed in immature neurons, mechanisms for Cl− import may be

required

0 10 20 30 40 50 60 70 80 90 100

-140

-120

-100

-80

-60

-40

-20

0

[Cl-]

i (mM)

EC

l (m

V)

Resting membrane potential (Vm)

ECl

in mature neurons

ECl

in immature neurons

7

3) When ECl is equal or close to Vm, opening of GABAARs may not cause

any significant change in Vm. The open GABAARs, however, may counteract and

prevent depolarization that would otherwise be caused by simultaneous excitatory Na+

influx. Such an effect of the GABA-activated conductance is known as "shunting

inhibition". It is important to note, however, that also during shunting inhibition,

despite that no change in voltage is seen, GABAAR-activation is associated with Cl−

flux that may impose a considerable load on the postsynaptic neuron (Doyon et al.,

2016b), and thus a need for Cl− extrusion mechanisms.

Cation chloride cotransporters (CCCs)

In the CNS, a family of cation chloride cotransporters (CCCs), encoded by the genes

Slc12a 1-9, plays a central role in controlling [Cl–]i. The CCCs are membrane

glycoproteins that mediate ion transport across the plasma membrane (Blaesse et al.,

2009). The family is broadly classified into four categories;

1) Na+-Cl–-cotransporter (NCC)

2) Na+-K+ -Cl–-cotransporters (NKCC1-2)

3) K+-Cl–-cotransporters (KCC1-4)

4) Cation chloride cotransporter 9 (CCC9) and CCC-interacting protein 1

(CIP1).

Regarding neuronal Cl− regulation, NKCC1 and KCC2 are highly important because

they are widely expressed in neurons throughout the CNS.

KCC2: Neurons express four different isoforms of K+-Cl–-cotransporters (KCCs)

KCC1, KCC2, KCC3 and KCC4. KCCs are secondary active transporters which do

not consume energy directly, but rather use the K+ gradient as an energy source to

extrude Cl– (Alvarez-Leefmans & Delpire, 2009). KCC2 is an electroneutral neuron-

specific Cl– transporter responsible for maintaining a low [Cl–]i. In contrast to other

KCCs, which are activated by osmotic changes, KCC2 is also active in isotonic

conditions (Acton et al., 2012). This makes KCC2 the main Cl– extruder that enables

neuronal inhibition mediated by hyperpolarizing influx of Cl– via glycine and GABAA

receptors. KCC2 exists in two splice variants: KCC2a and KCC2b. KCC2a expression

is relatively low but stable throughout the lifespan whereas expression of KCC2b

increases during development (Uvarov et al., 2007). KCC2 is widely expressed in

central neurons with predominant localization in the somato-dendritic compartments.

The compartmentalization of KCC2 often results in local differences in [Cl–] and

EGABA, for instance in soma as compared with dendrites (Raimondo et al., 2012).

8

The secondary and tertiary structures of KCC2 and all other CCC family members are

yet to be resolved. However, the structure predicted on basis of hydrophobicity

suggests that KCC2 has 12 transmembrane domains, intracellularly located N- and C-

termini and short loops connecting the transmembrane domains (Payne et al., 1996;

Gamba, 2005; Hartmann & Nothwang, 2014) (Figure 3). The C- and N-termini

contain several amino acid residues (serine, threonine and tyrosine; Figure 3) that may

be selectively and specifically de/phosphorylated by several phosphatases/kinases

(Kahle et al., 2013). This occurs in many different physiological and

pathophysiological conditions, dynamically affecting KCC2 activity and thereby local

[Cl–]i distribution (for review Kelsch et al., 2001; Kahle et al., 2005; Lee et al., 2007).

The C-terminal of KCC2 also contains an ISO domain which is necessary for KCC2

activity in isotonic conditions (Acton et al., 2012).

Figure 3. Putative structure of the KCC2 protein, with amino acid residues important for

transport activity and membrane stability indicated. Phosphorylation, mutation or deletion of

these amino acid residues result in changes of ion-transport or surface stability. (Reproduced

from Medina et al., 2014, with permission).

KCC2 may form di- and oligomeric complexes in the membrane, but the functional

significance of KCC2 oligomerization remains unclear as experimental evidence on

the oligomerization-dependent transport activity of KCC2 is still sparse and highly

contradictory (Medina et al., 2014). Age-dependent upregulation of KCC2 expression

mediates a developmental switch of the mode of GABAAR signaling in the CNS

9

(Ganguly et al., 2001; Tyzio et al., 2008). Early in development, the relatively low

level of KCC2 function in immature neurons keeps the Cl– gradient inefficient for

driving hyperpolarizing influx of Cl–. Instead, activation of GABAAR then usually

leads to depolarizing efflux of Cl– and an excitatory action of GABA. With

development, upregulation of KCC2 leads to a much lower [Cl–]i, compatible with Cl–

influx and enabling the GABA-evoked inhibition typical for the mature CNS.

Perturbed KCC2 activity may result in altered GABAAR-mediated signaling, as

documented in stress (Hewitt et al., 2009; Bains et al., 2015; MacKenzie & Maguire,

2015) and in pathological conditions, including epilepsy (Kaila & Miles, 2010;

Khirug et al., 2010; Kaila, et al., 2014b; Karlócai et al., 2015), chronic pain (Coull et

al., 2003; De Koninck, 2009; Gagnon et al., 2013), autism (Lemonnier & Ben-Ari,

2010; Tyzio et al., 2014) and spinal cord injury (Boulenguez et al., 2010).

Figure 4. Main types of neuronal Cl– transport. KCC2-mediated Cl– extrusion is driven by the

K+ gradient generated by the Na+-K+-ATPase. Extrusion of Cl– generates an inwardly directed

electrochemical gradient for Cl– resulting in hyperpolarizing currents through Cl– permeable

channels (top). NKCC1-mediated Cl– uptake is driven by the Na+ gradient generated by the

Na+-K+-ATPase. Uptake of Cl– by NKCC1 generates an outwardly directed electrochemical

gradient for Cl– resulting in depolarizing currents through Cl– permeable channels (bottom).

NKCC1: Na+-K+-Cl– cotransporter 1 (NKCC1) is highly expressed in immature

neurons, promoting intracellular Cl– accumulation and thereby depolarizing GABAAR

10

responses (Delpire, 2000). NKCC1 uses the Na+ gradient as energy source to transport

K+ and Cl– into the cell and it mediates electroneutral transport with 1Na+:1K+:2Cl–

stoichiometry (Alvarez-Leefmans & Delpire, 2009). The expression of NKCC1 is

downregulated in the first three weeks of postnatal development (Li et al., 2002). In

mature neurons, the distribution of NKCC1 is poorly studied but recent results

indicate that NKCC1 is highly expressed in the axon initial segment where the action

of GABA may be depolarizing even in mature neurons (Khirug et al., 2008). In

contrast to mature central neurons, primary sensory (Delpire & Austin, 2010) and

olfactory neurons (Kaneko et al., 2004) maintain a high [Cl–]i and exhibit depolarizing

GABA responses as a consequence of a high level of NKCC1. Figure 4 shows the

main types of Cl– transport by KCC2 and NKCC1.

Thermodynamics of KCC2 and NKCC1

KCC2 mediates electroneutral transport of 1K+:1Cl–. This means that the overall

transport of K+ and Cl– neither generates a net current across the membrane nor

depends on the membrane potential. Thus, the driving force (UKCC2) for KCC2-

mediated K+ and Cl– transport solely depends on the chemical potential differences

for K+ and Cl–:

UKCC2 = RT ln([Cl–]i/[ Cl–]o) + RT ln([K+]i/[K+]o)

At equilibrium UKCC2 = 0 and thus:

RT ln([Cl–]i/[ Cl–]o) + RT ln([K+]i/[K+]o) = 0

[Cl–]i/[Cl–]o = [K+]o/[K+]i

[Cl–]i = [Cl–]o[K+]o/[K+]i (2)

If [K+]i is set by the Na+-K+-ATPase or by experimental solutions, then equation (2)

provides the [Cl–]i at KCC2 equilibrium. With the experimental values used in this

study ([Cl–]o = 146.4 mM, [K+]o = 5 mM and [K+]i = 140 mM), [Cl–]i will be 5.2 mM,

and thus, according to the Nernst equation, ECl = –83 mV. This is below the average

Vm, implying that GABAAR activation will usually hyperpolarize the cell.

In the absence of other mechanisms than KCC2 for Cl– transport, steady-state [Cl–]i

(and ECl) can thus be obtained from equation (2). Then, ECl = EK. However, all neurons

likely display at least some gCl, which will drive ECl towards the resting (or prevailing)

potential. This also means that if KCC2 is blocked, [Cl–]i will redistribute passively

and at equilibrium ECl = Vm. With KCC2 present, as we have expressed in paper IV,

steady-state ECl will settle somewhere between EK and resting Vm, depending on the

11

relative contribution of KCC2 and gCl. Papers III and IV of this thesis clarify the

influence of gCl on the Cl– homeostasis.

Since KCC2 functions near thermodynamic equilibrium (Payne, 1997; Payne et al.,

2003), relatively small perturbations in Cl– or K+ concentrations may affect the

transport markedly. Thus, for instance, even a moderate increase in [K+]o, such as

often detected in epileptic conditions (Viitanen et al., 2010), may reverse Cl− transport

by KCC2 to become inwardly directed (Payne, 1997; Thompson & Gähwiler, 1989).

Similarly as KCC2, NKCC1 mediates electroneutral transport. Thus, the equilibrium

for NKCC1 depends on the chemical gradients of Na+, K+ and Cl– and at equilibrium:

[Cl–]i = [Cl–]o ([Na+]o [K+]o / ([Na+]i [K+]i))1/2

However, in contrast to KCC2 and as a consequence of the physiological ion

concentrations, NKCC1 usually operates far from equilibrium and mediates Cl–

transport in inward direction only (Alvarez-Leefmans & Delpire, 2009).

Role of impermeant anions in setting [Cl–]i

Neurons are surrounded by a complex network of extracellular matrix (ECM). The

ECM is mostly composed of chondroitin sulphate proteoglycans (CSPGs) and

hyaluronic acid (Dityatev et al., 2010; Soleman et al., 2013). The CSPGs contain very

many fixed negative charges. Similarly, the cell cytoplasm contains negatively

charged macromolecules, such as nucleic acids and proteins, which cannot pass the

plasma membrane. It has recently been proposed that the negatively charged

macromolecules on the in- and outside of the cell membrane play a critical role in

setting the resting [Cl–]i (Glykys et al., 2014a). This idea, which contrasts with the

view that [Cl–]i is determined by the CCCs (Kaila et al., 2014a; Rivera et al., 1999),

rests on the knowledge that impermeant ions may affect the distribution of permeant

ions via a so called Gibbs-Donnan effect (see e.g. Kutchai, 2004, for a pedagogical

description). In support of their proposal, Glykys et al. (2014a) reported that digestion

of the ECM was associated with a rise in [Cl–]i. The report by Glykys et al. (2014a),

which also stated that CCCs do not determine resting [Cl–]i, has been intensely

debated (Ben-Ari, 2014; Cho, 2014; Delpire & Staley, 2014; Glykys et al., 2014b;

Luhmann et al., 2014; Voipio et al., 2014; Vogt, 2015; Doyon et al., 2016b; Robel &

Sontheimer, 2016). It therefore remains unclear if the ECM does play a critical role

in setting the resting [Cl–]i. This question was addressed in paper II of the present

thesis.

12

Specific aims of the thesis

The overall aim of the present thesis is to clarify the mechanisms involved in the

neuronal chloride homeostasis, which is critically important for the fast neuronal

inhibition in the central nervous system. The more specific aims are:

1. To clarify how to accurately estimate the intracellular Cl– concentration,

[Cl–]i, by electrophysiological methods.

2. To clarify the role of the extracellular matrix (ECM) in setting [Cl–]i.

3. To develop a method for estimating the resting Cl– conductance of the

neuronal membrane.

4. To elucidate the role of passive, conductive Cl– flux for the neuronal [Cl–]i

homeostasis.

5. To establish the theoretical relation between the Cl– equilibrium potential

(ECl), Cl– extrusion capacity and synaptic input.

13

Methodological considerations

In the present thesis, the regulation of neuronal [Cl–]i was investigated using

electrophysiological, biochemical and computational techniques. A brief summary of

the methods is provided below and the detailed experimental procedures are presented

in the methods section of each original publication and manuscript.

Cell and tissue preparation

The present work was made using neurons from the preoptic hypothalamic region of

Sprague-Dawley rats. The main reason for the choice of preparation is that we have a

large amount of available information on the electrophysiological signaling properties

of these neurons. It is known that GABA signaling undergoes a developmental

transition from excitation to inhibition. Since this transition is completed during the

first two weeks of postnatal development (Tyzio et al., 2008), we used 3-5 weeks old

rats to ensure that GABA signaling is inhibitory. Older rats were not used since in

general, neurons form younger rats survive better and are more easily accessed with

the techniques used.

Experiments were mainly performed on neurons from the medial preoptic nucleus

(MPN). These neurons express functional GABAARs and GlyRs (Karlsson et al.,

1997a) and as such, they provide an excellent experimental model for studying [Cl–]i

regulation. One advantage of cells with two types of Cl–-permeable channels is that

we may estimate [Cl–]i by using an agonist to one receptor/channel type while

simultaneously blocking the other type to understand its role for the [Cl–]i

homeostasis.

In this study, two acute in vitro preparations were used: 1) slice preparation 2) acutely

dissociated neurons.

Brain slices, 200-300 µm thick and containing the preoptic area were prepared as

described in detail in the Methods section of paper I-III (see also e. g. Johansson et

al., 1995; Karlsson et al., 2011)) The major advantage of the slice preparation is the

near-physiological conditions, with intact neurites, synaptic connections, surrounding

glial cells and ECM. However, neurons in slice preparation preserve extensive

dendritic branches which makes it impossible to control the distal membrane voltage

by the voltage-clamp technique.

An enzyme-free, mechanical dissociation method was used to dissociate neurons from

brain slices as described previously by Vorobjev (1991) with the modifications

according to Karlsson et al. (1997b). To dissociate neurons, a vibrating glass rod was

applied to the site of the MPN. Dissociated neurons lack the majority of neurites,

14

enabling excellent control of the membrane potential in voltage-clamp conditions.

They also allow rapid application of pharmacological drugs. In addition, this

preparation preserves many functional presynaptic nerve terminals and their

properties can be tested in this preparation. However, since neurites and natural

environment is lost, a comparison with neurons in slice preparations is often

warranted.

Electrophysiological recording

The present work was mainly carried out by using patch-clamp techniques. Patch-

clamp was first introduced by Neher & Sakmann (1976). Since then it has been widely

used in studying ionic transport mediated through ion channels and transporter

proteins. This study was conducted using different patch-clamp recording

configurations, viz conventional whole-cell recording (Hamill et al., 1981),

amphotericin (Rae et al., 1991) or gramicidin (Abe et al., 1994; Ebihara, et al., 1995;

Kyrozis & Reichling, 1995) perforated-patch recording and cell-attached recording

(Hamill et al., 1981; Alcami et al., 2012).

To carry out patch-clamp recordings, a borosilicate-glass pipette with a tip of ~1 µm

dimeter was produced in a standard electrode puller. A patch pipette filled with

pipette-filling solution, usually mimicking the natural intra- or extracellular ionic

composition, and with a resistance of 3.5- 4.5 MΩ was approached to the cell. When

the pipette touched the cell, weak suction was applied to form a tight seal of resistance

usually > 1 GΩ. To achieve an electrical continuity with the interior of the cell, the

membrane was ruptured by applying a pulse of negative pressure. This conventional

whole-cell configuration was used in the experiments to dialyze the cytoplasm of the

cell with intracellular pipette-filling solution. Despite its simplicity, conventional

whole-cell configuration has limitations, which include run-down of currents over

time due to washout of essential cytoplasmic components (Pusch & Neher, 1988). To

avoid this problem, either amphotericin- or gramicidin-perforated patch-clamp

technique was used. Amphotericin forms non-selective monovalent ion pores in the

patch membrane and allows electrical access to the cell interior (Rae et al., 1991). To

achieve an amphotericin-perforated patch, amphotericin was added to the intracellular

solution in the pipette and a GΩ seal between pipette and cell membrane was formed.

Amphotericin pores are permeable to Cl– ions, implying a disturbed [Cl–]i via

Cl– exchange between cytosol and recording pipette. Thus, to study unperturbed

[Cl–]i, the gramicidin-perforated patch technique was used. Since the pores formed by

gramicidin in the cell membrane are impermeable to Cl–, [Cl–]i is not affected by flux

into or out of the recording pipette (Ebihara et al., 1995).

In slice preparations, spontaneous neuronal activity was recorded either in the cell-

attached configuration or with perforated-patch recording. We also stimulated the

15

presynaptic fibers within the MPN and recorded the evoked and spontaneous

postsynaptic responses as described previously by Malinina et al. (2006). The current

recorded in the cell-attached configuration during neuronal action potential generation

is composed of a capacitive component as well as a resistive component. The resistive

component has a time course similar to the transmembrane potential, while the

capacitive component has a time course that corresponds to the derivative of that

potential (Johansson & Århem, 1992).

During all electrophysiological recordings, neurons were continuously perfused with

extracellular solution via a gravity-fed perfusion system. Exchange to solution

containing GABA, glycine and/or other substances was controlled by solenoid valves

via a computer. The majority of electrophysiological studies were performed at room

temperature, 21-23 C. (Exceptions are noted in paper II.)

Electrophysiological measuring of [Cl–]i

[Cl–]i, was obtained from the Cl– equilibrium potential, ECl, as given by the current-

voltage (I-V) relation for the current component carried exclusively by Cl–. First, the

I-V relation for “leak” currents was recorded in the absence of ligands for Cl–

permeable channels. The I-V relation was obtained from a voltage ramp at a rate of

±1.6 V s-1 (Karlsson et al., 2011). To obtain the current carried by Cl–, leak currents

were subtracted from the current measured in the presence of GABA or glycine.

[Cl–]i was estimated from the known concentration of extracellular Cl– ([Cl–]o) and

ECl, as defined by the Nernst equation. Measuring [Cl–]i requires activation of Cl–

permeable channels that, depending on DFCl, may affect [Cl–]i during the

measurement. Thus, to minimize artificial changes, [Cl–]i was always measured at a

holding membrane potential close to the reversal potential (ECl), where Cl– flux is

small. Further, a technique with symmetric ramps in negative and positive direction

was used to avoid artificial [Cl–]i perturbations. That this technique gives negligible

perturbation of [Cl–]i was shown theoretically in Fig. 2F,G of paper I.

Electrophysiological measurement of [Cl–]i provides some clear advantages over

optical Cl– imaging. Electrophysiological techniques are usually called “invasive”

since contact between pipette and cytoplasm is achieved. However, as noted above,

the gramicidin-perforated patch technique enables recordings that do not perturb

[Cl–]i. Further, ECl provides the probably most direct and sensitive method available

for quantifying [Cl–]i. “Non-invasive” optical imaging requires artificial introduction

of fluorescent Cl–-binding sensors into the cell and is far from reliable in estimating

[Cl–]i (Berglund et al., 2011). The main two limitations of Cl– imaging to measure

[Cl–]i are:

16

1) Changes in pH are commonly observed during neuronal excitation and all

available Cl– sensors have a high sensitivity to H+ ions. Therefore, measuring

[Cl–]i highly depends on changes in pH.

2) Commonly used Cl– sensors have a high Kd (~ 90 mM) and low Cl–

sensitivity. In physiological conditions, changes in [Cl–]i are often in a range

of a few mM. This implies that the sensitivity of Cl– sensors in physiological

conditions is limited.

Cl– transporter activity and [Cl–]i homeostasis

The neuronal [Cl–]i homeostasis involves two major mechanisms of Cl– transport: one

being mediated by CCCs and the other one being passive and mediated by the flux of

Cl– via available channels. CCCs mediate electroneutral ion transport and therefore,

an electrophysiological approach to directly monitor the ion flux mediated by CCCs

is not possible. However, the activity of CCCs can be estimated by monitoring

changes in [Cl–]i. To study the [Cl–]i homeostasis, cells were loaded to a high [Cl–]i,

by a combination of membrane depolarization and activation of Cl– channels by either

GABA or glycine. Depolarization generates a driving force that drives Cl– into the

cell. During [Cl–]i loading as well as during recovery, [Cl–]i was estimated as

previously described and plotted against time. The time constant of [Cl–]i recovery

was obtained from a single-exponential fit. To study Cl– transporter-mediated [Cl–]i

changes during recovery, contribution of passive channel-mediated leak was

minimized by keeping the command voltage close to ECl. Pharmacological blockers

were applied to discriminate different transporters involved in the [Cl–]i recovery.

To study the contribution of Cl– leak conductance in [Cl–]i recovery, cells were loaded

to a high [Cl–]i and during the recovery phase, the membrane was voltage clamped

close to the resting membrane potential (in this case -74 mM) and transporter proteins

were blocked. This provided the essential driving force for Cl– leak-mediated [Cl–]i

recovery.

Estimating relative Cl– conductance by a novel method

In paper III, we have developed a novel electrophysiological method for measuring

relative resting Cl– conductance (gCl). We have formulated a simple mathematical

relation which is based on the conductance version of the Goldman-Hodgkin-Katz

equation (Goldman, 1943; Hodgkin & Katz, 1949). We showed that the relative

resting gCl can be obtained from the change in resting membrane potential (ΔErev) and

the change in Cl– equilibrium potential (ΔECl) under conditions where [Cl–]i is altered.

[Cl–]i-dependent changes in Erev and ECl were estimated by using electrophysiological

techniques. This method depends on the simple, and experimentally verified,

assumption that the total membrane conductance does not change before and after Cl–

17

loading. To minimize the effect of other factors, the experiments were performed in

presence of a cocktail of blockers: Tetraethylammonium and tetrodotoxin were used

to block unwanted gating of voltage-dependent K+ and Na+ channels, Cd2+ was used

to block presynaptic-release and VU0255011‐1 was used to block KCC2 activity.

In principle, the method for estimating gCl may be used for all preparations accessible

to common electrophysiological recording techniques. In cases where [Cl–]i cannot

easily be manipulated, [Cl–]o can be changed instead. Then, offset changes in liquid-

junction potentials and electrode potentials must also be taken into account.

Immunoblotting

Whole-cell KCC2 protein from MPN-containing coronal slices was extracted using

RIPA buffer containing a protease inhibitor. Membrane-expressed proteins were

extracted using a plasma membrane protein extraction kit. Following extraction,

proteins in individual samples were measured and separated by sodium dodecyl

sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred from the gel

to a polyvinylidene difluoride (PVDF) membrane. Then, the PVDF membranes were

incubated with the primary antibodies against KCC2, transferrin receptor, calnexin or

β-tubulin. The membranes were subsequently incubated with the appropriate HRP-

conjugated secondary antibody and incubated with western HRP chemiluminescence

substrate. The chemiluminescent signals in the membranes were acquired using an

Odyssey Fc Imaging System and quantified using Image studio version 3.1. Western

blot experiments were repeated at least 8-10 times on independent samples (animals).

For analysis, all signals from the membrane fraction were normalized to transferrin

whereas those from the whole-cell fraction were normalized to β-tubulin or calnexin.

Obtained data were compared using non-parametrical tests.

Computational modeling

In this thesis, computational modeling was used to either complement

electrophysiological data or to address a specific question that is difficult to

understand using electrophysiological experiments only.

In paper I, a computational model of GABA-activated currents was used to clarify the

changes in [Cl–]i that are expected from the Cl– current under voltage-clamp

conditions. The model was based on that made by Karlsson et al. (2011) and the

kinetics of GABA-mediated currents observed by Haage et al. (2005), but here we

omitted the desensitization. Initially, the model parameter values were adopted to the

experimental conditions used by Bianchi & Macdonald (2002), to achieve roughly

similar currents as observed experimentally by those authors. Subsequently, we

18

computed the Cl– current and the changes in [Cl–]i. Further, we used this model to

complement our electrophysiological data obtained from MPN neurons.

In paper III, the computational model described by Karlsson et al. (2011) was used to

study the influence of resting gCl on the [Cl–]i homeostasis. Thus, the model was

modified to account for resting gCl as well as for KCC2-mediated transport. The

KCC2-mediated transport capacity was expressed in terms of a proportionality

constant (gKCC2) that may be seen as an “equivalent conductance” although the

cotransport with K+ makes Cl– transport by KCC2 electroneutral. The constants gKCC2

and gCl were first adjusted to match experimental [Cl–]i changes with KCC2-mediated

or conductive Cl– transport in isolation. Subsequently, the time course of theoretically

expected [Cl–]i recovery at various combinations of gKCC2, gCl and membrane potential

was computed.

In paper IV, to clarify how [Cl–]i depends on Cl– extrusion capacity and synaptic input,

we formulated, in mathematical terms, the relation between steady-state ECl, gKCC2,

total gCl and Vm and illustrated this in graphical form. The basic idea behind the

interpretation in terms of steady synaptic input is that synaptic inhibition will mainly

affect gCl (as for GABA- or glycine-gated channels) whereas synaptic excitation,

which is most often due to non-selective (e. g. glutamate-gated) cation channels, will

be equivalent to a positive change of Vm.

19

Results and discussion

Project 1: Estimating [Cl–]i by electrophysiological methods

To study the Cl– homeostasis, it is crucial that an appropriate method for measuring

Cl– dynamics is selected. The method should be accurate and sufficiently sensitive to

detect small changes in [Cl–]i. Several electrophysiological or imaging techniques are

available to measure [Cl–]i in neurons. Common electrophysiological methods are

based on an estimate of ECl which is used together with the known extracellular Cl–

concentration, [Cl–]o, to calculate [Cl–]i from the Nernst equation. ECl is measured as

the voltage of zero current (reversal potential) under conditions where the current is

carried by Cl– only. In principle, ECl reflects [Cl–]i directly and if ECl can be measured

with mV-precision, [Cl–]i can usually, for physiological levels, be obtained with at

least mM-precision. (As noted above, on the contrary, recently developed, non-

invasive Cl– sensors have a low Cl– sensitivity in the physiological range of [Cl–]i

(Berglund et al., 2011)). When estimating ECl on basis of membrane currents, care

must be taken to separate the current component carried by Cl– from other current

components. In the present work, GABA and glycine were used to induce currents

carried by Cl–, since the studied cells express Cl–-permeable ligand-gated ion channels

(GABAA and glycine receptors) sensitive to these substances. Although these

channels also show some permeability for HCO3–, the contribution of HCO3

– was

excluded by using a HCO3–-free extracellular solution. Further, leak current

components were quantified in the absence of ligands and subtracted from the total

membrane current.

To estimate ECl, the I-V relation of the studied Cl– currents must be established. The

two most common techniques, here called the interpolation and the voltage-ramp

methods, differ with respect to the voltage protocol used. Since the introduction of the

voltage-clamp technique (Hodgkin & Huxley, 1952), the interpolation method has

been the most widely used. For studies of ligand-gated channels, it usually implies

that the ligand is applied repeatedly, each time at a different steady voltage. The I-V

relation for a specific time point after the start of ligand application (or at the time of

peak current) is subsequently obtained by interpolation of the current amplitudes

recorded at each voltage.

In the voltage-ramp method, on the other hand, voltage is increased or decreased

gradually at a steady rate while continuously recording the current in the presence of

a ligand. Recently, by using the voltage ramp method Karlsson et al. (2011) showed

that desensitization of GABAARs is significantly slower than previously reported and

that changes in [Cl–]i largely define the decay of GABA-evoked current. These

findings contradict previous ideas that GABAARs desensitize (i. e. channels close at

prolonged GABA exposure) rapidly and that [Cl–]i does not change much during a

20

typical experiment with activation of GABA- or glycine activated channels. Earlier

conclusions were, however, to a large extent based on analysis with the interpolation

method. For instance, in a previous study, Bianchi & Macdonald (2002) concluded,

on basis of analysis with the interpolation method, that in roughly similar conditions

as in the study by Karlsson et al. (2011), the decay of GABA-evoked current is mainly

due to receptor desensitization and changes in [Cl–]i are negligible. The discrepancy

between conclusions based on the two methods, led us to the hypothesis that the

interpolation and the voltage-ramp methods differ with respect to how correctly

[Cl–]i and gCl are described by the two methods. To clarify which method is the most

correct one and to what degree the two methods agree with theoretical expectations

of [Cl–]i and gCl, we made a direct comparison between the two methods.

To clarify the theoretically expected changes in ECl and gCl, we used a computational

model, based on that described by Karlsson et al. (2011), to compare the GABA-

evoked currents expected with the two types of voltage protocol. For simplicity, we

assumed that the studied channels showed no true desensitization, that is, the channels

stay open in the continued presence of GABA. The model was adopted to the

conditions for GABA-evoked currents as observed by Bianchi & Macdonald (2002).

Most essentially, the model accounted for the changes in [Cl–]i that would be expected

from the Cl– flux through the GABA-activated channels. For simplicity, we assumed

that no other channels or transporters contributed to changes in [Cl–]i.

The numerically computed currents at different steady holding voltages showed a

substantial decay during the activation of the GABA-mediated conductance. The I-V

relations constructed from current amplitude at four different time points showed a

roughly constant interpolated reversal potential, which if taken to reflect ECl suggests

that there is no change in [Cl–]i. Further, the slope of the I-V relation decreased with

time and, if taken to reflect gCl, suggests a substantial desensitization. However, since

the model did not include any channel desensitization, the current decay must have

another cause. One advantage with the model, in comparison with wet experiments,

is that [Cl–]i can be obtained directly from the computations. Here, the model clearly

demonstrated that [Cl–]i changed dramatically during the currents recorded at steady

voltages. Thus, in essence, the computations show that the I-V relation obtained by

the interpolation method gives a reversal potential that does not correctly reflect ECl

and [Cl–]i and a slope that does not correctly reflect gCl, contrary to common

assumptions, such as made by Bianchi & Macdonald (2002). On the other hand,

computations of the currents and [Cl–]i during voltage ramps showed that I-V relations

obtained from these ramps correctly reflected the changes in [Cl–]i and also correctly

reflected the constant gCl of the model. The model also displayed that although the

altered Cl– flux induced by the voltage ramps to some degree changes [Cl−]i, the use

of symmetrical ramps in negative and positive direction minimizes such changes to

non-significant levels.

21

Wet experiments with dissociated MPN neurons confirmed the differences between

the two protocols to estimate [Cl–]i and gCl. Since the biological GABAARs show some

degree of true desensitization, the slope of the I-V curves from the living neurons

decay, although slowly, with time in the presence of GABA.

The difference between the two methods for constructing I-V plots depend on the

interpolation used in the non-accurate method. As explained in Fig. 5 of paper I, this

implies that artificial I-V relations (colored in the figure) are obtained whereas the

true I-V relations (black and gray in the figure) remain undetected.

The principle that flux through ion channels may be associated with changes in ion

concentrations apply to all type of channels and need to be considered in every

experiment involving currents through ion channels. Our findings suggest that the

voltage-ramp method is to be preferred when I-V relations reflecting [Cl–]i and gCl are

needed. This also means that a number of earlier studies may need reinterpretation.

The study by Bianchi & Macdonald (2002) provides a particularly clear example.

However, the interpolation method has been used even in the first papers where the

voltage-clamp technique was introduced (Hodgkin & Huxley, 1952). This should not

be taken to imply that all such studies are mistaken. In many cases, for instance most

likely for the fast Na+ channels studied by Hodgkin & Huxley (1952), the magnitude

and time course of ion flux are so limited with respect to the volumes affected that

significant changes in ion concentrations may not take place. In case of channels with

slow kinetics, such as GABAA- and glycine receptors and several types of voltage-

gated K+ channels, however, interpretation of the results from the interpolation

method should be made with caution.

22

Project 2: Role of Cl– transporters and extracellular matrix in [Cl–]i

homeostasis

The effect of GABAAR-meditated signaling depends critically on the transmembrane

Cl− electrochemical gradient. It has been widely accepted that cation-chloride

cotransporters (CCCs) play a primary role in setting [Cl–]i and the transmembrane

Cl− electrochemical gradient. Recently, Glykys et al. (2014a) used a genetically

expressed Cl– sensor to measure the resting [Cl–]i in young hippocampal neurons.

Surprisingly, the authors concluded that CCCs are not involved in setting resting

[Cl–]i. Furthermore, they observed an increased [Cl–]i upon digesting the negatively

charged proteoglycans in the extracellular matrix (ECM). Thus, the authors claimed

that [Cl–] is not actively maintained by CCCs, but rather passively distributed

according to a Gibbs-Donnan equilibrium under the influence of fixed local anionic

charges (Glykys et al., 2014a). The mechanism proposed by Glykys et al. (2014a) has

been intensely debated (Ben-Ari, 2014; Cho, 2014; Delpire & Staley, 2014; Glykys

et al., 2014b; Luhmann et al., 2014; Voipio et al., 2014; Silayeva et al., 2015; Vogt,

2015; Doyon et al., 2016b; Robel & Sontheimer, 2016).

Paper II of the present thesis is an attempt to at least partly resolve the debated role of

the ECM and of CCCs for [Cl–]i and synaptic transmission, using electrophysiological

measurements in a slice preparation containing MPN neurons. [Cl–]i, was estimated

on the basis of gramicidin-perforated recordings and I-V relations constructed from

voltage ramps as described in paper I. First, we quantified resting [Cl–]i, which was

significantly lower than expected from a thermodynamic equilibrium (or passive

distribution). This clearly contradicts the findings of Glykys et al. (2014a) while

indicating (likely indirectly active) Cl– extrusion involved in setting [Cl–]i. Another

controversial finding of Glykys et al. (2014a) was that inhibiting KCC2 or NKCC1

activity had no effect on ECl, directly contradicting the previously assumed role of

CCCs for [Cl–]i (Rivera et al., 1999; Kaila et al., 2014a) and also contradicting their

own earlier findings (Dzhala & Staley, 2003; Dzhala et al., 2005, 2008; Glykys et al.,

2009). However, when we, in the present work, blocked KCC2 activity by a KCC2-

specific blocker, [Cl–]i increased significantly. These results clearly show that [Cl–]i

is maintained below thermodynamic equilibrium by the activity of KCC2.

As noted above, Glykys et al. (2014a) observed an increased [Cl–]i upon digesting the

negatively charged proteoglycans (chondroitin sulfate) in the ECM. Surprisingly, in

our experimental conditions too, degradation of the ECM with chondroitinase

increased basal [Cl–]i. Recently, Delpire & Staley (2014), in support of the findings

by Glykys et al. (2014a), proposed that digestion of local impermeant anionic charges

increases local [Cl–]o (thus differing from [Cl–]o in the bulk solution) thereby changing

the driving force for Cl– (but see Kaila et al., 2014a). The analysis in the present work,

showed that the [Cl–]o contributing to the driving force for GABA-evoked currents,

23

was not affected by ECM degradation and was, when Cl– leak was minimized, not

significantly different from bulk [Cl–]o. This contradicts the idea that digestion of local

impermeant anionic charges affects the local [Cl–]o.

Our results thus showed that although the ECM affects [Cl–]i in MPN neurons, this is

not likely via effects of the macromolecular anionic charges. Since our results also

demonstrated that [Cl–]i is dependent on KCC2 activity, we hypothesized that the

action of ECM digestion may be mediated by effects on KCC2. Such effects would

be consistent with the dynamic control of KCC2 expression by several factors, e.g.

neuronal activity (Rivera et al., 2004; Khirug et al., 2010; Puskarjov et al., 2012),

changes in cell volume (Kahle et al., 2005; Friedel et al., 2015; Jentsch, 2016),

neurotrophic factors (Rivera et al., 2004; Boulenguez et al., 2010), interaction with

other molecules (Gauvain et al., 2011; Ivakine et al., 2013; Mahadevan et al., 2014)

and [Cl–]i (Piala et al., 2014). We tested our hypothesis by studying KCC2 activity as

reflected by the recovery after loading to a high [Cl–]i. (We here used the voltage-

ramp method analyzed in paper I.) The [Cl–]i recovery showed an exponential time

course and was clearly dependent on KCC2. The slowing of [Cl–]i recovery upon

digesting the ECM thus suggests that KCC2 is affected.

Alterations in KCC2 protein level have been associated with several physiological

and pathological conditions (Kahle et al., 2013). Therefore, we asked whether the

suggested effect of ECM digestion on KCC2 function was due to an altered KCC2

protein level. KCC2 protein is present in monomeric, dimeric and oligomeric forms

(Blaesse et al., 2006; Watanabe et al., 2009), but the importance of monomer/oligomer

KCC2 remains debatable (Jaenisch et al., 2010; Medina et al., 2014; Leonzino et al.,

2016). We found that ECM digestion reduced the amount of KCC2 monomers, in

whole-cell as well as plasma membrane fractions, but did not affect KCC2 dimers.

The parallell reduction of KCC2 monomers and Cl– extrusion capacity, suggests that

monomers are likely to be functional for Cl– transport.

We also investigated the mechanism underlying the KCC2 downregulation. KCC2

can be downregulated by the Ca2+-dependent protease calpain (Puskarjov et al., 2012)

and via an NMDA-receptor-dependent pathway (Lee et al., 2011). In our samples,

blocking NMDA receptors did not inhibit the ChABC-dependent downregulation of

KCC2. At the same time, in presence of the calpain-inhibitor calpeptin (Puskarjov et

al., 2012) or high-threshold Ca2+-channel blocker Cd2+ (Sundgren-Andersson &

Johansson, 1998), ChABC failed to reduce the KCC2 level, suggesting that ECM

removal triggers influx of Ca2+ that mediates calpain-dependent cleavage of KCC2.

Potentially, different types of Ca2+ channels could mediate Ca2+ entry (Takahashi &

Momiyama, 1993). Previously, Kochlamazashvili et al. (2010) reported that

enzymatic removal of another component of ECM, hyaluronic acid, reduces L-type

voltage-dependent Ca2+ currents and had profound impact on long-term potentiation.

24

Our western blot data in paper II, however, suggests that Ca2+ influx through N-type

Ca2+ channels contribute to the KCC2 downregulation upon ECM degradation. This

is consistent with the idea that digestion of ECM leads to activation/potentiation of

N-type Ca2+ channels and thereby downregulation of KCC2. However, further

experiments are warranted to unravel the mechanism behind the ECM-dependent

modulation of N-type Ca2+ channels.

Recently, Doyon et al. (2016a) showed by computer simulations that mild alteration

in KCC2 activity may cause incorrect neuronal coding via altered GABAAR-mediated

transmission. In paper II, we tested directly and experimentally the hypothesis that

ChABC-dependent downregulation of KCC2 affects the efficacy of GABAAR-

mediated transmission, in the MPN-containing slice preparation. The preparation is

suitable for studying inhibition, since even in the absence of excitatory glutamatergic

inputs, many MPN neurons are spontaneously active. Thus, presynaptic stimulation

readily induce inhibition seen as a clear reduction of postsynaptic impulse frequency.

Since our results, at low stimulus frequency, showed a clear inhibition also in ChABC-

treated brain slices, the effect of ECM digestion on KCC2, [Cl–]i and ECl apparently

was not sufficient to prevent the inhibitory action of GABA, although it may have

been mediated by shunting rather than hyperpolarization (cf Khazipov et al., 2014).

Intense neuronal activity may, however, accentuate accumulation of internal Cl–

(Staley et al., 1995). In agreement, we found that high-frequency presynaptic

stimulation was followed by significantly reduced GABAAR-mediated inhibition or

even by excitation in ChABC-treated cells, but not in control cells. Thus, intense

neuronal activity after ECM digestion may change a dominant inhibitory GABA

action to excitatory. In summary, paper II highlights a critical role of the ECM for

GABAAR-mediated signaling. Although digestion of the ECM increases [Cl–]i, the

mechanism is not compatible with that proposed by Glykys et al. (2014a). Rather,

digestion of the ECM decreases the expression and function of KCC2, consequently

causing an increase in resting [Cl–]i as well as an impaired capacity to handle a high

Cl– load.

The ECM has an important role in development (Bonnans et al., 2014), plasticity

(Kerrisk et al., 2014) and in the pathogenesis of injuries and disease (Fawcett, 2015).

Degradation of the ECM (Lendvai et al., 2000; Stern et al., 2001; Trachtenberg &

Stryker, 2001) and reduction of neuronal inhibition (Harauzov et al., 2010) is critical

for experience-dependent plasticity. Thus, our findings may provide the underlying

mechanism of experience-dependent changes in neuronal excitability. These results

also offer novel insights into understanding pathological conditions, such as epilepsy,

spinal cord injury and chronic pain, where the ECM may be degraded (Wilczynski et

al., 2008; Soleman et al., 2013) and the effect on [Cl–]i may result in increased

neuronal activity (Coull et al., 2003; Coull et al., 2005; Boulenguez et al., 2010;

25

Nardou et al., 2011). Further experiments are needed to differentiate the role of

different ECM components in physiological and pathological conditions.

26

Project 3: Influence of Cl– leak conductance on [Cl–]i homeostasis

According to the predominant view, the resting neuronal membrane shows some

conductance for K+, Na+ and Cl– ions (Goldman, 1943; Hodgkin & Katz, 1949;

Forsythe & Redman, 1988). The resting leak (or background) conductance is

important for setting the resting membrane potential as well as for the control of

neuronal excitability, and ionic homeostasis (Hille, 2001). Although it has long been

known that the neuronal membrane possesses some resting Cl– conductance

(Goldman, 1943; Hodgkin & Katz, 1949; Forsythe & Redman, 1988), data on gCl in

central neurons is extremely scarce. First, the quantity of resting gCl is not well known.

Secondly, in spite of a number of characterized Cl– channel types (Jentsch et al.,

2002), the molecular identity of Cl– leak channels is unknown. Recently, Rinke et al.

(2010) proposed that ClC-2, a voltage-dependent Cl– channel, may mediate a major

part of the leak conductance. The single-channel conductance of ClC-2 channels is

~2-3 pS and the presence of these channels influences [Cl–]i in hippocampal neurons.

However, the expression of ClC-2 is highly restricted to specific neuronal types

(Smith et al., 1995) and even restricted to specific neuronal compartments (Földy et

al., 2010). On the other hand, in central neurons, GABAARs are the main mediators

of Cl– conductance and they are widely expressed in the brain. It has been known that

extrasynaptic GABAARs are non-desensitizing, high-affinity receptors (Belelli &

Lambert, 2005; Kullmann et al., 2005) that may mediate a large leak Cl– conductance

(single-channel conductance ~30 pS) (Wagner, 2005). In paper III, an attempt was

made to clarify the molecular identity of Cl– leak channels in MPN neurons. Further,

we studied the influence of Cl– leak conductances on the [Cl–]i homeostasis. We have

also developed a novel method to estimate the relative Cl– leak conductance and used

it to obtain as estimate of resting gCl in MPN neurons.

It has been proposed that inwardly rectifying ClC-2 channels control [Cl–]i by acting

as one-way Cl– exit valves or Cl– efflux pathways (Staley, 1994; Földy et al., 2010;

Rinke et al., 2010; Smart, 2010). Contrary to that, Ratté & Prescott (2011) claim that

ClC-2 cannot acts as an efflux pathway because of its imperfect inward rectification.

Rather, ClC-2s act as Cl– leak channels. To clarify the role of Cl– leak (or passive)

conductances, their influence must be separated from that of transporters involved in

[Cl–]i regulation. Therefore, we first studied [Cl–]i recovery as described previously

(paper II), in the presence of a KCC2- or NKCC1-blocker. The KCC2 blocker

significantly slowed the [Cl–]i recovery, whereas the NKCC1 blocker did not affect

the time course of [Cl–]i recovery. Further, we did not observe ClC-2 mediated

currents in the MPN neurons. These results indicate that [Cl–]i recovery after a high

load is mainly mediated by KCC2 while NKCC1 and ClC-2 are not involved. In paper

III, we also showed that even with KCC2 blocked, [Cl–]i recovers from a high Cl– load

if the voltage is kept near the resting potential, but not at a depolarized level. These

results clearly demonstrate that Cl– extrusion depends on voltage as expected for Cl–

27

leak channels, where voltage contribute to the driving force for Cl– (cf equation 1).

The level of [Cl–]i will, in the absence of transporter proteins, be dictated by the

voltage, with steady ECl settling at the membrane potential.

GABAARs are widely expressed in the CNS where they mediate a Cl– conductance in

several different conditions. Spontaneous GABA release and tonic activation of

GABAARs may contribute to a more or less steady Cl– conductance (Farrant &

Nusser, 2005). We therefore tested whether GABAARs contribute to the conductance

underlying the passive [Cl–]i extrusion. However, neither a GlyR blocker nor a

competitive GABAAR antagonist affected conductive [Cl–]i recovery. This indicates

that [Cl–]i recovery does not depend on activation by glycine or GABA of their

respective receptor types. However, the open-channel GABAAR blocker picrotoxin

reduced conductive [Cl–]i recovery. Our interpretation is that spontaneously open

GABAARs, such as shown by Birnir et al. (2000) and Wlodarczyk et al. (2013), may

contribute to the gCl important for conductive [Cl–]i recovery. It is known that a low

concentration of ambient GABA activates tonic GABAARs which are primarily

expressed in extrasynaptic sites (Semyanov et al., 2003) and that δ-subunit containing

GABAARs are commonly expressed at such sites (Gunn et al., 2015). We here showed

that pharmacological activation of δ-subunit containing GABAARs dramatically

potentiated the conductive [Cl–]i recovery. With respect to the role of increased [Cl–]i

for epileptic conditions (Khalilov et al., 2003), this may be taken to support the idea

that enhancing tonic inhibition may help in preventing epilepsy. In addition, there is

a number of neurological conditions where perturbed signaling by δ-subunit

containing GABAARs has been reported, such as e.g. mood disorders (Maguire et al.,

2005; Maguire & Mody, 2008; Sequeira et al., 2009), schizophrenia (Maldonado-

Avilés et al., 2009) and fragile X syndrome (Curia et al., 2009) for which our findings

may be helpful in better understanding the etiology of disease.

Neurosteroids are endogenous substances shown to have anti-seizure effects and may

be important in pathological conditions like stress and epilepsy. A high concentration

(20 nM - 100 nM) of the endogenous GABAAR-active neurosteroid allopregnanolone

has been detected in the brain (Purdy et al., 1991; Mareš et al., 2006; Biagini et al.,

2010). Therefore, we studied whether allopregnanolone-mediated potentiation of

synaptic GABAARs affects the [Cl–]i recovery. Our results show that spontaneous

GABA release does not contribute to conductive [Cl–]i recovery in basal conditions.

However, allopregnanolone enhanced the speed of recovery. Since the effect was

observed already at low steroid concentration shown to affect GABAARs activated by

GABA and with impulse generation blocked, this was interpreted as due to enhanced

effects of spontaneously released GABA (cf Haage & Johansson, 1999). At higher

concentration, known to directly activate GABAARs in MPN neurons (Haage &

Johansson, 1999), allopregnanolone had a dramatic effect on conductive [Cl–]i

28

recovery. Thus our results indicate that if Vm is sufficiently hyperpolarized, then the

neurosteroid has the capacity to strongly enhance [Cl–]i recovery.

In this project we developed a novel, but simple, method for estimating the resting

gCl. Such a method is needed since previous attempts to estimate gCl were associated

with considerable problems and depended on many assumptions (cf Forsythe &

Redman, 1988). We derived a simple mathematical expression where the relative gCl

is obtained from the experimentally accessible shift in reversal potential of total

current (corresponding to the resting potential) relative to the shift in ECl under

conditions where [Cl–]i is altered. The method depends on the assumption that the

total membrane conductance does not change with Cl– loading, a condition that was

experimentally verified. The method should be relatively easy to apply to many other

preparations. An attractive alternative to changing [Cl–]i by Cl– loading, as done here,

may be to simply alter [Cl–]o. If that is done, care has to be taken to account for

changes in liquid-junction and electrode potentials that accompany changes in [Cl–]o.

With this method, we showed that in MPN neurons, gCl accounts for ~ 6 % of the total

resting conductance. Here the measured resting gCl represents the relative Cl– leak

conductance in the cell body of the MPN neuron. However, it is possible that the Cl–

leak conductance is compartment specific, so that dendrites or neurons with intact

dendrites have different relative Cl– leak conductance. The majority of the Cl– leak

conductance was blocked by picrotoxin, an open channel blocker of GABAARs,

suggesting that GABAARs account for a major fraction of resting gCl. Alterations in

synaptic and extrasynaptic GABAAR subunits and expression have been reported in

many neurological disorders, such as epilepsy (Stell et al., 2003), chronic pain (Iura

et al., 2016), autism (Fernandez et al., 2007; Braudeau et al., 2011), stress (Gauthier

& Nuss, 2015; for excellent review, see Hines et al., 2012). Such alterations should

be associated with altered Cl– leak conductance. Therefore, our method likely paves

the way for new approaches in understanding the role of an altered Cl– leak

conductance and thereby neuronal excitability in physiological and pathological

conditions.

Our electrophysiological experimental results clearly indicate that MPN neurons

express a resting gCl that influences the recovery of [Cl–]i. To complement the

experimental results, we used a computational model with parameters matching the

experimental values. The reason for this study was to clarify the role of conductive

Cl– flux in [Cl–]i homeostasis. It was clear that at negative potentials, conductive Cl–

flux may account for a substantial [Cl–]i recovery, with a time course roughly similar

to that of KCC2 dependent recovery. Interestingly, our simulations showed that at a

negative resting potential, the computed [Cl–]i recovery with gCl as well as KCC2-

mediated transport was significantly quicker than with KCC2 alone. Taken together,

our findings offer new insights into understanding [Cl–]i homeostasis and normal

inhibition via the control of the Cl– leak conductance.

29

Dissociated MPN neurons had a roughly exponential Cl– recovery with a time

constant ~500 s. The obtained time constant was dramatically slower than the 82 s

observed in slice preparations (paper II) and the few seconds reported in some earlier

studies (Staley & Proctor, 1999; Wagner et al., 2001). The difference is likely, at least

partly, due to recording techniques, since previous measurement of [Cl–]i was

performed either with sharp-electrodes or by whole-cell patch clamp which dialyze

the cell and cause ionic equilibration with the patch pipette solution. As shown by

Pusch & Neher (1988), the latter equilibration may occur within a few seconds. Thus,

the quicker [Cl–]i recovery in studies with sharp-electrodes or by whole-cell patch

clamp may be artefactual. The different time course seen in dissociated cells as

compared with slice preparations may possibly be attributed to either difference in

surface-to-volume ratio in these preparations or absence of ECM in the former

preparations. As a consequence of greater number of remaining neurites, the surface-

to-volume ratio should be higher for neurons in slice preparations, and it seems

possible that they may express more KCC2 than dissociated neurons. In slice

preparations, neurons are surrounded by the ECM, whereas dissociated neurons lack

the integrity of the ECM which is crucial for expression of KCC2 (paper II). It seems

possible that dissociation of individual neurons from slices may damage the ECM and

downregulate KCC2 which may lead to slower [Cl–]i recovery.

30

Project 4: Clarifying the role of KCC2 under conditions with different

inhibitory or excitatory conductances

Neurons constantly receive inhibitory and excitatory signals that work by altering the

ion conductance of the membrane. The efficiency of GABAAR-mediated inhibition

relies heavily on ECl which is dependent on the clearance of Cl– from the intracellular

cytosol by KCC2 (Doyon et al., 2016b). There is a large body of evidence showing

that, in pathological conditions, downregulation of KCC2 activity leads to altered

[Cl–]i homeostasis (Price et al., 2005). Recently, Doyon et al. (2016a) emphasized that

the effects on [Cl–]i by even small changes in KCC2 activity may have dramatic

consequences for neuronal coding. As previously described (paper III), the membrane

Cl– conductance, gCl, can also profoundly influence the [Cl–]i homeostasis. To

understand the control of [Cl–]i, it is important not only to separately clarify the

transport capacity by KCC2 and the influence of gCl, but also how these factors

interact to determine ECl and [Cl–]i. Since the effect of gCl (but not that of KCC2)

depends on the membrane potential, Vm, also this has to be considered as a

determinant of [Cl–]i. In paper IV, we formulated the relation between the above

factors in mathematical terms, to enable a graphical description and better

understanding of their influence on ECl. We thus showed that in a manner that strongly

depends on gCl, ECl will vary between the equilibrium potential for K+ ions, EK, and

Vm (The dependence on EK is due to the co-transport of Cl– with K+ by KCC2). Our

finding that the influence of the transport capacity by KCC2 (expressed as gKCC2) is

reduced with increased gCl, is in stark contrast with the recent claim by Doyon et al.

(2016b) that ECl is sensitive to changes in KCC2 activity only when gCl is large. Doyon

et al. (2016b) used this idea to explain the controversial findings of Glykys et al.

(2014a) who did not observe changes in the resting [Cl–]i upon blocking KCC2.

Doyon et al. (2016b) proposed that the findings of Glykys et al. (2014a) could be

explained by their use of a slice preparation that lacked strong inhibitory synaptic

activity. They concluded that “Testing under low Cl– load conditions will, therefore,

lead to gross underestimation of the contribution of KCC2 and overestimation of the

contribution of the Donnan effect” (Doyon et al., 2016b). However, the relation

between gCl and transport capacity was not correctly formulated by Doyon et al.

(2016b). As we showed in paper IV, at low steady (inhibitory or resting) gCl, [Cl–]i

will strongly depend on KCC2 activity, whereas at high steady gCl, [Cl–]i will depend

more on the prevailing membrane potential. In contrast, a lack of excitatory synaptic

input, which affects [Cl–]i mainly via Vm, will reduce the dependence of [Cl–]i on

KCC2 and could in principle contribute to the controversial findings of Glykys et al.

(2014a).

31

Overall Summary

In the first project (paper I), we compared the most common methods to study ion

channel properties and the associated interpolation and voltage-ramp methods for

constructing I-V relations. The two methods clearly generate different results: the

popular interpolation method may fail to detect conductance and ion concentration

changes during prolonged channel activation, whereas the voltage-ramp method

accurately detects such changes.

In the second project (paper II), we clarified the role of the ECM for resting [Cl–]i and

Cl– transport capacity. Our results showed that digestion of the ECM increases [Cl–]i.

In parallel, digestion of the ECM decreases the expression and function of KCC2, thus

explaining the increase in resting [Cl–]i. Further, ECM degradation together with

intense neuronal activity may change the effect of GABA-mediated signaling from

inhibitory to excitatory. We also clarified severeal steps in the ECM-dependent

signaling pathway that leads to the downregulation of KCC2.

In the third project (paper III), we clarified the influence of the Cl– leak conductance

on [Cl–]i homeostasis in MPN neurons. We first developed a novel method to estimate

the relative Cl– leak conductance and obtained a measure of about 6 % in MPN

neurons. We also showed that, at resting potential, the Cl– leak conductance

contributes significantly, and in the same order of magnitude as KCC2, to the recovery

from a raised [Cl–]i. We identified spontaneously open GABAARs as the major

molecular contributors to resting Cl– leak conductance and conductive [Cl–]i recovery.

We also showed that an endogenous neurosteroid as well as a synthetic GABA-active

substance may enhance conductive [Cl–]i recovery.

In the fourth project (paper IV), we formulated the mathematical relationship between

ECl, gCl, Vm and KCC2 transport activity at steady state. We showed that [Cl–]i strongly

depends on KCC2 transport capacity when gCl is low. We also showed that, in the

absence of other transporters than KCC2, ECl will settle between Vm and EK,

depending on the relative magnitude of gCl and KCC2 transport capacity.

32

Acknowledgements

First and foremost, I would like to express my sincere thanks to my supervisor Prof.

Staffan Johansson for scientific and personal mentorship. You have been an inspiring

teacher and a scientifically- rigorous mentor. I remain indebted to you for all that I

learnt in the lab under your constant guidance.

I also would like to thank my co-supervisor Dr. Micheal Druzin for introducing me to

patch-clamp. I can’t thank you enough for your scientific guidance, constructive

comments and suggestions.

I also owe many thanks to past and present members of the lab especially Evgenya,

Tatiana, Erii, Olga and Nazanin.

Thank you Anita Dreyer-Perkiomaki for the timely administrative help.

I also thank Paul Kingham and Mallappa for your help and advice.

I would also like to thank Vahid, Cecilia, Stylianous, Paolo and all colleagues at IMB

for enriching the working atmosphere.

Edward, Sebastian and Vaishnavi- I owe you many thanks for your help, support and

friendship.

Joydeep, Prashant, Shrikant, Pratibha, Dilip, Anjana, Amol, Sonali, Chinmay,

Chaitanya, Dharmesh, Edvin, Brijesh and Manoj- you all have been amazing friends.

Thanks for making my stay in Umeå memorable.

I would also like to thank Abhilasha and my family for your unconditional love,

support and patience.

33

References

Abe, Y., Furukawa, K., Itoyama, Y. & Akaike, N. (1994). Glycine response in acutely

dissociated ventromedial hypothalamic neuron of the rat: new approach with

gramicidin perforated patch-clamp technique. Journal of Neurophysiology 72, 1530–

1537.

Acton, B.A., Mahadevan, V., Mercado, A., Uvarov, P., Ding, Y., Pressey, J., … & Woodin,

M.A. (2012). Hyperpolarizing GABAergic transmission requires the KCC2 C-terminal

ISO domain. Journal of Neuroscience 32, 8746-8751.

Alcami, P., Franconville, R., Llano, I. & Marty, A. (2012). Measuring the firing rate of high-

resistance neurons with cell-attached recording. Journal of Neuroscience, 32, 3118–

3130.

Alvarez-Leefmans, F.J. & Delpire, E. (2009). Thermodynamics and kinetics of chloride

transport in neurons: An outline. In Physiology and pathology of chloride transporters

and channels in the nervous system, Alvarez-Leefmans, F.J. & Delpire, E. (eds), pp 81-

108, Academic Press, London.

Bains, J.S., Cusulin, J.I.W. & Inoue, W. (2015). Stress-related synaptic plasticity in the

hypothalamus. Nature Reviews Neuroscience 16, 377–388.

Barbaccia, M.L., Serra, M., Purdy, R.H. & Biggio, G. (2001). Stress and neuroactive steroids.

International Review of Neurobiology 46, 243–272.

Belelli, D. & Lambert, J.J. (2005). Neurosteroids: endogenous regulators of the GABAA

receptor. Nature Reviews Neuroscience 6, 565–575.

Ben-Ari, Y. (2002). Excitatory actions of gaba during development: the nature of the nurture.

Nature Reviews Neuroscience 3, 728–739.

Ben-Ari, Y. (2014). The GABA excitatory/inhibitory developmental sequence: a personal

journey. Neuroscience 279, 187–219.

Ben-Ari, Y., Cherubini, E., Corradetti, R. & Gaiarsa, J.-L.L. (1989). Giant synaptic potentials

in immature rat CA3 hippocampal neurones. Journal of Physiology 416, 303–325.

Berglund, K., Kuner, T., Feng, G. & Augustine, G.J. (2011). Imaging synaptic inhibition with

the genetically encoded chloride indicator Clomeleon. Cold Spring Harbor Protocols

2011, 1492–1497.

Biagini, G., Panuccio, G. & Avoli, M. (2010). Neurosteroids and epilepsy. Current Opinion

in Neurology 23, 170–176.

Bianchi, M.T. & Macdonald, R.L. (2002). Slow phases of GABAA receptor desensitization:

structural determinants and possible relevance for synaptic function. Journal of

Physiology 544, 3–18.

Birnir, B., Everitt, A.B., Lim, M.S. & Gage, P.W. (2000). Spontaneously opening GABAA

channels in CA1 pyramidal neurones of rat hippocampus. Journal of Membrane

Biology 174, 21–29.

Blaesse, P., Airaksinen, M.S., Rivera, C. & Kaila, K. (2009). Cation-chloride cotransporters

and neuronal function. Neuron 61, 820–838.

Blaesse, P., Guillemin, I., Schindler, J., Schweizer, M., Delpire, E., Khiroug, L., … &

Nothwang, H.G. (2006). Oligomerization of KCC2 correlates with development of

inhibitory neurotransmission. Journal of Neuroscience 26, 10407–10419.

Bonnans, C., Chou, J. & Werb, Z. (2014). Remodelling the extracellular matrix in

34

development and disease. Nature Reviews Molecular Cell Biology 15, 786–801.

Bormann, J., Hamill, O.P. & Sakmann, B. (1987). Mechanism of anion permeation through

channels gated by glycine and gamma-aminobutyric acid in mouse cultured spinal

neurones. Journal of Physiology 385, 243–286.

Boron, W.F. (2004). Regulation of intracellular pH. Advances in Physiology Education 28,

160–179.

Boron, W.F. & Russell, J.M. (1983). Stoichiometry and ion dependencies of the intracellular-

pH-regulating mechanism in squid giant axons. Journal of General Physiology 81,

373–399.

Boulenguez, P., Liabeuf, S., Bos, R., Bras, H., Jean-Xavier, C., Brocard, C., … & Vinay, L.

(2010). Down-regulation of the potassium-chloride cotransporter KCC2 contributes to

spasticity after spinal cord injury. Nature Medicine 16, 302–307.

Braudeau, J., Delatour, B., Duchon, A., Pereira, P.L., Dauphinot, L., de Chaumont, F., … &

Potier, M.-C. (2011). Specific targeting of the GABAA receptor α5 subtype by a

selective inverse agonist restores cognitive deficits in Down syndrome mice. Journal of

Psychopharmacology 25, 1030–1042.

Brickley, S.G., Cull-Candy, S.G. & Farrant, M. (1996). Development of a tonic form of

synaptic inhibition in rat cerebellar granule cells resulting from persistent activation of

GABAA receptors. Journal of Physiology 497, 753–759.

Bäckström, T. (1976). Epileptic seizures in women related to plasma estrogen and

progesterone during the menstrual cycle. Acta Neurologica Scandinavica 54, 321–347.

Cho, C.H. (2014). Star players sidelined in chloride homeostasis in neurons. Frontiers in

Cellular Neuroscience 8, 114. doi.org/10.3389/fncel.2014.00114

Concas, A., Mostallino, M.C., Porcu, P., Follesa, P., Barbaccia, M.L., Trabucchi, M., … &

Biggio, G. (1998). Role of brain allopregnanolone in the plasticity of gamma-

aminobutyric acid type A receptor in rat brain during pregnancy and after delivery.

Proceedings of the National Academy of Sciences of the United States of America 95,

13284–13289.

Corpéchot, C., Robel, P., Axelson, M., Sjövall, J. & Baulieu, E.E. (1981). Characterization

and measurement of dehydroepiandrosterone sulfate in rat brain. Proceedings of the

National Academy of Sciences of the United States of America 78, 4704–4707.

Coull, J.A.M., Boudreau, D., Bachand, K., Prescott, S.A., Nault, F., Sík, A., … & De

Koninck, Y. (2003). Trans-synaptic shift in anion gradient in spinal lamina I neurons as

a mechanism of neuropathic pain. Nature 424, 938–942.

Coull, J.A.M., Beggs, S., Boudreau, D., Boivin, D., Tsuda, M., Inoue, K., … & De Koninck,

Y. (2005). BDNF from microglia causes the shift in neuronal anion gradient underlying

neuropathic pain. Nature 438, 1017–1021.

Curia, G., Papouin, T., Séguéla, P. & Avoli, M. (2009). Downregulation of tonic GABAergic

inhibition in a mouse model of fragile X syndrome. Cerebral Cortex 19, 1515–1520.

De Koninck, Y. (2010). Modulation of chloride homeostasis by microglia. In Physiology and

pathology of chloride transporters and channels in the nervous system, Alvarez-

Leefmans, F.J. & Delpire, E. (eds), pp 471-488, Academic Press, London.

Delpire, E. (2000). Cation-chloride cotransporters in neuronal communication. News in

Physiological Sciences 15, 309–312.

Delpire, E. & Austin, T.M. (2010). Kinase regulation of Na+-K+-2Cl– cotransport in primary

afferent neurons. Journal of Physiology 588, 3365–3373.

35

Delpire, E. & Staley, K.J. (2014). Novel determinants of the neuronal Cl– concentration.

Journal of Physiology 592, 4099–4114.

Dityatev, A., Schachner, M. & Sonderegger, P. (2010). The dual role of the extracellular

matrix in synaptic plasticity and homeostasis. Nature Reviews Neuroscience 11, 735–

746.

Doyon, N., Prescott, S.A. & De Koninck, Y. (2016a). Mild KCC2 hypofunction causes

inconspicuous chloride dysregulation that degrades neural coding. Frontiers in

Cellular Neuroscience 9. doi.org/10.3389/fncel.2015.00516.

Doyon, N., Vinay, L., Prescott, S.A. & De Koninck, Y. (2016b). Chloride regulation: A

dynamic equilibrium crucial for synaptic inhibition. Neuron 89, 1157–1172.

Dzhala, V.I., Brumback, A.C. & Staley, K.J. (2008). Bumetanide enhances phenobarbital

efficacy in a neonatal seizure model. Annals of Neurology 63, 222–235.

Dzhala, V.I. & Staley, K.J. (2003). Excitatory actions of endogenously released GABA

contribute to initiation of ictal epileptiform activity in the developing hippocampus.

Journal of Neuroscience 23, 1840–1846.

Dzhala, V.I., Talos, D.M., Sdrulla, D.A., Brumback, A.C., Mathews, G.C., Benke, T.A., … &

Staley, K.J. (2005). NKCC1 transporter facilitates seizures in the developing brain.

Nature Medicine 11, 1205–1213.

Ebihara, S., Shirato, K., Harata, N. & Akaike, N. (1995). Gramicidin-perforated patch

recording: GABA response in mammalian neurones with intact intracellular chloride.

Journal of Physiology 484, 77–86.

Eccles, J., Nicoll, R.A., Oshima, T. & Rubia, F.J. (1977). The anionic permeability of the

inhibitory postsynaptic membrane of hippocampal pyramidal cells. Proceedings of the

Royal Society of London B: Biological Sciences 198, 345-361.

Farrant, M. & Kaila, K. (2007). The cellular, molecular and ionic basis of GABAA receptor

signalling. Progress in Brain Research 160, 59–87.

Farrant, M. & Nusser, Z. (2005). Variations on an inhibitory theme: phasic and tonic

activation of GABAA receptors. Nature Reviews Neuroscience 6, 215–229.

Fawcett, J.W. (2015). The extracellular matrix in plasticity and regeneration after CNS injury

and neurodegenerative disease. Progress in Brain Research 218, 213–226.

Fernandez, F., Morishita, W., Zuniga, E., Nguyen, J., Blank, M., Malenka, R.C. & Garner,

C.C. (2007). Pharmacotherapy for cognitive impairment in a mouse model of Down

syndrome. Nature Neuroscience 10, 411-413.

Forsythe, I.D. & Redman, S.J. (1988). The dependence of motoneurone membrane potential

on extracellular ion concentrations studied in isolated rat spinal cord. Journal of

Physiology 404, 83–99.

Friedel, P., Kahle, K.T., Zhang, J., Hertz, N., Pisella, L.I., Buhler, E., … & Medina, I. (2015).

WNK1-regulated inhibitory phosphorylation of the KCC2 cotransporter maintains the

depolarizing action of GABA in immature neurons. Science Signaling 8, ra65, doi:

10.1126/scisignal.aaa0354.

Földy, C., Lee, S.-H., Morgan, R.J. & Soltesz, I. (2010). Regulation of fast-spiking basket cell

synapses by the chloride channel ClC-2. Nature Neuroscience 13, 1047–1049.

Gagnon, M., Bergeron, M.J., Lavertu, G., Castonguay, A., Tripathy, S., Bonin, R.P., … & De

Koninck, Y. (2013). Chloride extrusion enhancers as novel therapeutics for

neurological diseases. Nature Medicine 19, 1524–1528.

Gamba, G. (2005). Molecular physiology and pathophysiology of electroneutral cation-

36

chloride cotransporters. Physiological Reviews 85, 423–493.

Ganguly, K., Schinder, A.F., Wong, S.T. & Poo, M.M. (2001). GABA itself promotes the

developmental switch of neuronal GABAergic responses from excitation to inhibition.

Cell 105, 521–532.

Gauthier, I. & Nuss, P. (2015). Anxiety disorders and GABA neurotransmission: a

disturbance of modulation. Neuropsychiatric Disease and Treatment 11, 165-175.

Gauvain, G., Chamma, I., Chevy, Q., Cabezas, C., Irinopoulou, T., Bodrug, N., … & Poncer,

J.C. (2011). The neuronal K-Cl cotransporter KCC2 influences postsynaptic AMPA

receptor content and lateral diffusion in dendritic spines. Proceedings of the National

Academy of Sciences of the United States of America 108, 15474–15479.

Glykys, J., Dzhala, V., Egawa, K., Balena, T., Saponjian, Y., Kuchibhotla, K.V., … & Staley,

K.J. (2014a). Local impermeant anions establish the neuronal chloride concentration.

Science 343, 670–675.

Glykys, J., Dzhala, V., Egawa, K., Balena, T., Saponjian, Y., Kuchibhotla, K.V., … & Staley,

K.J. (2014b). Response to Comments on “ Local impermeant anions establish the

neuronal chloride concentration .” Science 345, 1130.

Glykys, J., Dzhala, V. I., Kuchibhotla, K. V., Feng, G., Kuner, T., Augustine, G., … & Staley,

K.J. (2009). Differences in cortical versus subcortical GABAergic signaling: a

candidate mechanism of electroclinical uncoupling of neonatal seizures. Neuron 63,

657–672.

Goldman, D.E. (1943). Potential, impedance, and rectification in membranes. Journal of

General Physiology 27, 37–60.

Grosso, S., Luisi, S., Mostardini, R., Farnetani, M., Cobellis, L., Morgese, G., … & Petraglia,

F. (2003). Inter-ictal and post-ictal circulating levels of allopregnanolone, an

anticonvulsant metabolite of progesterone, in epileptic children. Epilepsy Research 54,

29–34.

Gunn, B.G., Cunningham, L., Mitchell, S.G., Swinny, J.D., Lambert, J.J. & Belelli, D. (2015).

GABAA receptor-acting neurosteroids: A role in the development and regulation of the

stress response. Frontiers in Neuroendocrinology 36, 28–48.

doi.org/10.1016/j.yfrne.2014.06.001.

Haage, D., Bäckström, T. & Johansson, S. (2005). Interaction between allopregnanolone and

pregnenolone sulfate in modulating GABA-mediated synaptic currents in neurons from

the rat medial preoptic nucleus. Brain Research 1033, 58–67.

Haage, D. & Johansson, S. (1999). Neurosteroid modulation of synaptic and GABA-evoked

currents in neurons from the rat medial preoptic nucleus. Journal of Neurophysiology

82, 143–51.

Hamill, O.P., Marty, A., Neher, E., Sakmann, B. & Sigworth, F.J. (1981). Improved patch-

clamp techniques for high-resolution current recording from cells and cell-free

membrane patches. Pflügers Archiv - European Journal of Physiology 391, 85–100.

Hammond, C. (2015). The ionotropic GABAA receptor. In Cellular and molecular

neurophysiology. 4th ed., Hammond, C., pp 199- 219, Academic Press, London.

Harauzov, A., Spolidoro, M., DiCristo, G., De Pasquale, R., Cancedda, L., Pizzorusso, T., …

& Maffei, L. (2010). Reducing intracortical inhibition in the adult visual cortex

promotes ocular dominance plasticity. Journal of Neuroscience 30, 361–71.

Hartmann, A.-M. & Nothwang, H.G. (2014). Molecular and evolutionary insights into the

structural organization of cation chloride cotransporters. Frontiers in Cellular

37

Neuroscience 8, 470. doi: 10.3389/fncel.2014.00470

Hartzell, H.C. (2009). Chloride channels: An historical perspective. In Physiology and

pathology of chloride transporters and channels in the nervous system, Alvarez-

Leefmans, F.J. & Delpire, E. (eds), pp 3-15, Academic Press, London.

Herd, M.B., Belelli, D. & Lambert, J.J. (2007). Neurosteroid modulation of synaptic and

extrasynaptic GABAA receptors. Pharmacology & Therapeutics 116, 20–34.

Hewitt, S.A., Wamsteeker, J.I., Kurz, E.U. & Bains, J.S. (2009). Altered chloride homeostasis

removes synaptic inhibitory constraint of the stress axis. Nature Neuroscience 12, 438–

443.

Hille, B. (2001). Ion channels of excitable membranes. 3rd ed., Sinauer, Sunderland.

Hines, R.M., Davies, P.A., Moss, S.J. & Maguire, J. (2012). Functional regulation of GABAA

receptors in nervous system pathologies. Current Opinion in Neurobiology 22, 552–

558.

Hodgkin, A.L. & Huxley, A.F. (1952). The dual effect of membrane potential on sodium

conductance in the giant axon of Loligo. Journal of Physiology 116, 497–506.

Hodgkin, A.L. & Katz, B. (1949). The effect of sodium ions on the electrical activity of giant

axon of the squid. Journal of Physiology 108, 37–77.

Iura, A., Takahashi, A., Hakata, S., Mashimo, T. & Fujino, Y. (2016). Reductions in tonic

GABAergic current in substantia gelatinosa neurons and GABAA receptor δ subunit

expression after chronic constriction injury of the sciatic nerve in mice. European

Journal of Pain 20, 1678–1688.

Ivakine, E.A., Acton, B.A., Mahadevan, V., Ormond, J., Tang, M., Pressey, J.C., … &

McInnes, R.R. (2013). Neto2 is a KCC2 interacting protein required for neuronal Cl–

regulation in hippocampal neurons. Proceedings of the National Academy of Sciences

of the United States of America 110, 3561–3566.

Jaenisch, N., Witte, O.W. & Frahm, C. (2010). Downregulation of potassium chloride

cotransporter KCC2 after transient focal cerebral ischemia. Stroke 41, e151-e159.

Jentsch, T.J. (2016). VRACs and other ion channels and transporters in the regulation of cell

volume and beyond. Nature Reviews Molecular Cell Biology 17, 293-307

Jentsch, T.J., Stein, V., Weinreich, F. & Zdebik, A.A. (2002). Molecular structure and

physiological function of chloride channels. Physiological Reviews 82, 503–568.

Johansson, S. & Århem, P. (1992). Impulses of variable amplitude recorded from cell-

attached patches on small hippocampal neurons. Brain Research 580, 185–188.

Johansson, S., Sundgren, A.K. & Klimenko, V. (1995). Graded action potentials generated by

neurons in rat hypothalamic slices. Brain Research 700, 240–244.

Jones, M.V. & Westbrook, G.L. (1995). Desensitized states prolong GABAA channel

responses to brief agonist pulses. Neuron 15, 181–191.

Kahle, K.T., Deeb, T.T.Z., Puskarjov, M., Silayeva, L., Liang, B., Kaila, K. & Moss, S.J.

(2013). Modulation of neuronal activity by phosphorylation of the K-Cl cotransporter

KCC2. Trends in Neurosciences 36, 726–737.

Kahle, K.T., Rinehart, J., de Los Heros, P., Louvi, A., Meade, P., Vazquez, N., … & Lifton,

R.P. (2005). WNK3 modulates transport of Cl– in and out of cells: implications for

control of cell volume and neuronal excitability. Proceedings of the National Academy

of Sciences of the United States of America 102, 16783–16788.

Kaila, K. & Miles, R. (2010). Chloride homeostasis and GABA signaling in temporal lobe

epilepsy. Epilepsia 51, Suppl. 5, 52.

38

Kaila, K., Price, T.J., Payne, J.A., Puskarjov, M.,& Voipio, J. (2014a). Cation-chloride

cotransporters in neuronal development, plasticity and disease. Nature Reviews

Neuroscience 15, 637–654.

Kaila, K., Ruusuvuori, E., Seja, P., Voipio, J. & Puskarjov, M. (2014b). GABA actions and

ionic plasticity in epilepsy. Current Opinion in Neurobiology 26, 34–41.

Kaila, K., Voipio, J., Paalasmaa, P., Pasternack, M. & Deisz, R.A. (1993). The role of

bicarbonate in GABAA receptor-mediated IPSPs of rat neocortical neurones. Journal of

Physiology 464, 273–289.

Kaneko, H., Putzier, I., Frings, S., Kaupp, U.B. & Gensch, T. (2004). Chloride accumulation

in mammalian olfactory sensory neurons. Journal of Neuroscience 24, 7931–7938.

Karlócai, M.R., Wittner, L., Tóth, K., Maglóczky, Z., Katarova, Z., Rásonyi, G., … &

Freund, T.F. (2015). Enhanced expression of potassium-chloride cotransporter KCC2

in human temporal lobe epilepsy. Brain Structure and Function 221, 3601-3615.

Karlsson, U., Druzin, M. & Johansson, S. (2011). Cl– concentration changes and

desensitization of GABAA and glycine receptors. Journal of General Physiology 138,

609–626.

Karlsson, U., Haage, D. & Johansson, S. (1997a). Currents evoked by GABA and glycine in

acutely dissociated neurons from the rat medial preoptic nucleus. Brain Research 770,

256-260.

Karlsson, U., Sundgren, A.K., Näsström, J. & Johansson, S. (1997b). Glutamate-evoked

currents in acutely dissociated neurons from the rat medial preoptic nucleus. Brain

Research 759, 270–276.

Kelsch, W., Hormuzdi, S., Straube, E., Lewen, A., Monyer, H. & Misgeld, U. (2001). Insulin-

like growth factor 1 and a cytosolic tyrosine kinase activate chloride outward transport

during maturation of hippocampal neurons. Journal of Neuroscience 21, 8339–8347.

Kerrisk, M.E., Cingolani, L.A. & Koleske, A.J. (2014). ECM receptors in neuronal structure,

synaptic plasticity, and behavior. Progress in Brain Research 214, 101-131.

Khalilov, I., Holmes, G.L. & Ben-Ari, Y. (2003). In vitro formation of a secondary

epileptogenic mirror focus by interhippocampal propagation of seizures. Nature

Neuroscience 6, 1079–1085.

Khazipov, R., Valeeva, G. & Khalilov, I. (2014). Depolarizing GABA and developmental

epilepsies. CNS Neuroscience & Therapeutics 21, 83–91.

Khirug, S., Ahmad, F., Puskarjov, M., Afzalov, R., Kaila, K. & Blaesse, P. (2010). A single

seizure episode leads to rapid functional activation of KCC2 in the neonatal rat

hippocampus. Journal of Neuroscience 30, 12028–12035.

Khirug, S., Yamada, J., Afzalov, R., Voipio, J., Khiroug, L. & Kaila, K. (2008). GABAergic

depolarization of the axon initial segment in cortical principal neurons is caused by the

Na-K-2Cl cotransporter NKCC1. Journal of Neuroscience 28, 4635–4639.

Kochlamazashvili, G., Henneberger, C., Bukalo, O., Dvoretskova, E., Senkov, O., Lievens, P.

M.-J., … & Dityatev, A. (2010). The extracellular matrix molecule hyaluronic acid

regulates hippocampal synaptic plasticity by modulating postsynaptic L-type Ca2+

channels. Neuron 67, 116–128.

Kullmann, D.M., Ruiz, A., Rusakov, D.M., Scott, R., Semyanov, A. & Walker, M.C. (2005).

Presynaptic, extrasynaptic and axonal GABAA receptors in the CNS: where and why?

Progress in Biophysics and Molecular Biology 87, 33–46.

39

Kutchai, H.C. (2004). Ionic equilibria and resting membrane potentials. In Physiology, 5th ed,

Berne, R.M., Levy, M.N, Koeppen, B.M. & Stanton, B.A. (eds), pp 22-30, Mosby,

Missouri.

Kyrozis, A. & Reichling, D.B. (1995). Perforated-patch recording with gramicidin avoids

artifactual changes in intracellular chloride concentration. Journal of Neuroscience

Methods 57, 27–35.

Lambert, J.J., Cooper, M.A., Simmons, R.D.J., Weir, C. J. & Belelli, D. (2009).

Neurosteroids: endogenous allosteric modulators of GABAA receptors.

Psychoneuroendocrinology 34, Suppl. 1, S48-S58.

Lee, H.H.C., Deeb, T.Z., Walker, J.A., Davies, P.A. & Moss, S.J. (2011). NMDA receptor

activity downregulates KCC2 resulting in depolarizing GABAA receptor-mediated

currents. Nature Neuroscience 14, 736–743.

Lee, H.H.C., Walker, J.A., Williams, J.R., Goodier, R.J., Payne, J.A. & Moss, S.J. (2007).

Direct protein kinase C-dependent phosphorylation regulates the cell surface stability

and activity of the potassium chloride cotransporter KCC2. Journal of Biological

Chemistry 282, 29777–29784.

Lemonnier, E. & Ben-Ari, Y. (2010). The diuretic bumetanide decreases autistic behaviour in

five infants treated during 3 months with no side effects. Acta Paediatrica,

International Journal of Paediatrics 99, 1885–1888.

Lendvai, B., Stern, E.A., Chen, B. & Svoboda, K. (2000). Experience-dependent plasticity of

dendritic spines in the developing rat barrel cortex in vivo. Nature, 404, 876–881.

Leonzino, M., Busnelli, M., Antonucci, F., Verderio, C., Mazzanti, M. & Chini, B. (2016).

The timing of the excitatory-to-inhibitory GABA switch is regulated by the oxytocin

receptor via KCC2. Cell Reports 15, 96–103.

Li, H., Tornberg, J., Kaila, K., Airaksinen, M.S. & Rivera, C. (2002). Patterns of cation-

chloride cotransporter expression during embryonic rodent CNS development.

European Journal of Neuroscience 16, 2358–2370.

Luhmann, H.J., Kirischuk, S. & Kilb, W. (2014). Comment on “Local impermeant anions

establish the neuronal chloride concentration”. Science 345, 1130.

Luisi, S., Petraglia, F., Benedetto, C., Nappi, R.E., Bernardi, F., Fadalti, M., … & Genazzani,

A.R. (2000). Serum allopregnanolone levels in pregnant women: Changes during

pregnancy, at delivery, and in hypertensive patients. Journal of Clinical Endocrinology

& Metabolism 85, 2429–2433.

MacAulay, N., Hamann, S. & Zeuthen, T. (2004). Water transport in the brain: Role of

cotransporters. Neuroscience 129, 1031–1044.

MacAulay, N. & Zeuthen, T. (2010). Water transport between CNS compartments:

Contributions of aquaporins and cotransporters. Neuroscience 168, 941–956.

Macdonald, R.L. & Botzolakis, E.J. (2009). GABAA receptor channels. In Physiology and

pathology of chloride transporters and channels in the nervous system, Alvarez-

Leefmans, F.J. & Delpire, E. (eds), pp 257-282, Academic Press, London.

MacKenzie, G. & Maguire, J. (2015). Chronic stress shifts the GABA reversal potential in the

hippocampus and increases seizure susceptibility. Epilepsy Research 109, 13–27.

Maguire, J. (2014). Stress-induced plasticity of GABAergic inhibition. Frontiers in Cellular

Neuroscience 8, 157. doi: 10.3389/fncel.2014.00157.

Maguire, J., Stell, B.M., Rafizadeh, M. & Mody, I. (2005). Ovarian cycle–linked changes in

GABAA receptors mediating tonic inhibition alter seizure susceptibility and anxiety.

40

Nature Neuroscience 8, 797–804.

Maguire, J. & Mody, I. (2008). GABAAR plasticity during pregnancy: Relevance to

postpartum depression. Neuron 59, 207–213.

Mahadevan, V., Pressey, J.C., Acton, B.A., Uvarov, P., Huang, M.Y., Chevrier, J., … &

Woodin, M.A. (2014). Kainate receptors coexist in a functional complex with KCC2

and regulate chloride homeostasis in hippocampal neurons. Cell Reports 7, 1762–1770.

Majewska, M.D., Harrison, N.L., Schwartz, R.D., Barker, J.L. & Paul, S.M. (1986). Steroid

hormone metabolites are barbiturate-like modulators of the GABA receptor. Science

232, 1004-1007.

Maldonado-Avilés, J.G., Curley, A.A., Hashimoto, T., Morrow, A.L., Ramsey, A.J.,

O’Donnell, P., … & Lewis, D.A. (2009). Altered markers of tonic inhibition in the

dorsolateral prefrontal cortex of subjects with schizophrenia. American Journal of

Psychiatry 166, 450–459.

Malinina, E., Druzin, M. & Johansson, S. (2006). Short-term plasticity in excitatory synapses

of the rat medial preoptic nucleus. Brain Research 1110, 128–135.

Mareš, P., Mikulecká, A., Haugvicová, R. & Kasal, A. (2006). Anticonvulsant action of

allopregnanolone in immature rats. Epilepsy Research 70, 110–117.

McDonald, B.J., Amato, A., Connolly, C.N., Benke, D., Moss, S.J. & Smart, T.G. (1998).

Adjacent phosphorylation sites on GABAA receptor β subunits determine regulation by

cAMP-dependent protein kinase. Nature Neuroscience 1, 23–28.

Medina, I., Friedel, P., Rivera, C., Kahle, K.T., Kourdougli, N., Uvarov, P. & Pellegrino, C.

(2014). Current view on the functional regulation of the neuronal K+-Cl– cotransporter

KCC2. Frontiers in Cellular Neuroscience 8, 27. doi.org/10.3389/fncel.2014.00027

Mellon, S.H. & Deschepper, C.F. (1993). Neurosteroid biosynthesis: genes for adrenal

steroidogenic enzymes are expressed in the brain. Brain Research 629, 283–92.

Mitchell, S.J. & Silver, R.A. (2003). Shunting inhibition modulates neuronal gain during

synaptic excitation. Neuron 38, 433–445.

Moss, S.J., Gorrie, G.H., Amato, A. & Smart, T.G. (1995). Modulation of GABAA receptors

by tyrosine phosphorylation. Nature 377, 344–348.

Moss, S.J. & Smart, T.G. (2001). Constructing inhibitory synapses. Nature Reviews

Neuroscience 2, 240–250.

Nardou, R., Yamamoto, S., Chazal, G., Bhar, A., Ferrand, N., Dulac, O., … & Khalilov, I.

(2011). Neuronal chloride accumulation and excitatory GABA underlie aggravation of

neonatal epileptiform activities by phenobarbital. Brain 134, 987–1002.

Neher, E. & Sakmann, B. (1976). Single-channel currents recorded from membrane of

denervated frog muscle fibres. Nature 260, 799–802.

Olsen, R.W. & Sieghart, W. (2008). International Union of Pharmacology. LXX. Subtypes of

gamma-aminobutyric acid A receptors: classification on the basis of subunit

composition, pharmacology, and function. Update. Pharmacological Reviews 60, 243–

260.

Olsen, R.W. & Sieghart, W. (2009). GABAA receptors: Subtypes provide diversity of

function and pharmacology. Neuropharmacology 56, 141–148.

Olsen, R.W. & Tobin, A.J. (1990). Molecular biology of GABAA receptors. FASEB Journal

4, 1469–1480.

Payne, J.A. (1997). Functional characterization of the neuronal-specific K-Cl cotransporter:

implications for [K+]o regulation. American Journal of Physiology 273, C1516–C1525.

41

Payne, J.A., Rivera, C., Voipio, J. & Kaila, K. (2003). Cation-chloride co-transporters in

neuronal communication, development and trauma. Trends in Neurosciences 26, 199–

206.

Payne, J.A., Stevenson, T.J. & Donaldson, L.F. (1996). Molecular Characterization of a

Putative K-Cl Cotransporter in Rat Brain: a neuronal-specific isoform. Journal of

Biological Chemistry 271, 16245–16252.

Piala, A.T., Moon, T.M., Akella, R., He, H., Cobb, M.H. & Goldsmith, E.J. (2014). Chloride

sensing by WNK1 involves inhibition of autophosphorylation. Science Signaling 7,

ra41. doi: 10.1126/scisignal.2005050.

Price, T.J., Cervero, F. & De Koninck, Y. (2005). Role of cation-chloride-cotransporters

(CCC) in pain and hyperalgesia. Current Topics in Medicinal Chemistry 5, 547–555.

Purdy, R.H., Morrow, A.L., Moore, P.H. & Paul, S.M. (1991). Stress-induced elevations of

gamma-aminobutyric acid type A receptor-active steroids in the rat brain. Proceedings

of the National Academy of Sciences of the United States of America 88, 4553–4557.

Pusch, M. & Neher, E. (1988). Rates of diffusional exchange between small cells and a

measuring patch pipette. Pflügers Archiv European Journal of Physiology 411, 204–

211.

Puskarjov, M., Ahmad, F., Kaila, K. & Blaesse, P. (2012). Activity-dependent cleavage of the

K-Cl cotransporter KCC2 mediated by calcium-activated protease calpain. Journal of

Neuroscience 32, 11356–11364.

Rae, J., Cooper, K., Gates, P. & Watsky, M. (1991). Low access resistance perforated patch

recordings using amphotericin B. Journal of Neuroscience Methods 37, 15–26.

Raimondo, J.V., Markram, H. & Akerman, C.J. (2012). Short-term ionic plasticity at

GABAergic synapses. Frontiers in Synaptic Neuroscience 4, 5.

doi.org/10.3389/fnsyn.2012.00005.

Ratté, S. & Prescott, S.A. (2011). ClC-2 channels regulate neuronal excitability, not

intracellular chloride levels. Journal of Neuroscience 31, 15838-15843.

Rinke, I., Artmann, J. & Stein, V. (2010). ClC-2 voltage-gated channels constitute part of the

background conductance and assist chloride extrusion. Journal of Neuroscience 30,

4776–4786.

Rivera, C., Voipio, J., Payne, J.A., Ruusuvuori, E., Lahtinen, H., Lamsa, K., … & Kaila, K.

(1999). The K+/Cl– co-transporter KCC2 renders GABA hyperpolarizing during

neuronal maturation. Nature 397, 251–255.

Rivera, C., Voipio, J., Thomas-Crusells, J., Li, H., Emri, Z., Sipila, S., … & Kaila, K. (2004).

Mechanism of activity-dependent downregulation of the neuron-specific K-Cl

cotransporter KCC2. Journal of Neuroscience 24, 4683–4691.

Robel, S. & Sontheimer, H. (2016). Glia as drivers of abnormal neuronal activity. Nature

Neuroscience 19, 28–33.

Ruffin, V. A., Salameh, A.I., Boron, W.F. & Parker, M.D. (2014). Intracellular pH regulation

by acid-base transporters in mammalian neurons. Frontiers in Physiology 5, 43. doi:

10.3389/fphys.2014.00043.

Sardini, A., Amey, J.S., Weylandt, K.H., Nobles, M., Valverde, M.A. & Higgins, C.F. (2003).

Cell volume regulation and swelling-activated chloride channels. Biochimica et

Biophysica Acta 1618, 153–162.

Semyanov, A., Walker, M.C. & Kullmann, D.M. (2003). GABA uptake regulates cortical

excitability via cell type–specific tonic inhibition. Nature Neuroscience 6, 484-90.

42

Semyanov, A., Walker, M.C., Kullmann, D.M. & Silver, R.A. (2004). Tonically active

GABAA receptors: modulating gain and maintaining the tone. Trends in Neurosciences

27, 262–269.

Sequeira, A., Mamdani, F., Ernst, C., Vawter, M.P., Bunney, W.E., Lebel, V., … & Turecki,

G. (2009). Global brain gene expression analysis links glutamatergic and GABAergic

alterations to suicide and major depression. PLoS One 4, e6585.

doi.org/10.1371/journal.pone.0006585.

Silayeva, L., Deeb, T.Z., Hines, R.M., Kelley, M.R., Munoz, M.B., Lee, H.H.C., … & Moss,

S.J. (2015). KCC2 activity is critical in limiting the onset and severity of status

epilepticus. Proceedings of the National Academy of Sciences of the United States of

America 112, 3523–3528.

Smart, T.G. (2010). Handling accumulated internal Cl– at inhibitory synapses. Nature

Neuroscience 13, 1043–1044.

Smith, R.L., Clayton, G.H., Wilcox, C.L., Escudero, K.W. & Staley, K.J. (1995). Differential

expression of an inwardly rectifying chloride conductance in rat brain neurons: a

potential mechanism for cell-specific modulation of postsynaptic inhibition. Journal of

Neuroscience 15, 4057–4067.

Soleman, S., Filippov, M.A., Dityatev, A. & Fawcett, J.W. (2013). Targeting the neural

extracellular matrix in neurological disorders. Neuroscience 253, 194–213.

Staley, K. (1994). The role of an inwardly rectifying chloride conductance in postsynaptic

inhibition. Journal of Neurophysiology 72, 273–284.

Staley, K.J. & Proctor, W.R. (1999). Modulation of mammalian dendritic GABAA receptor

function by the kinetics of Cl– and HCO3– transport. Journal of Physiology 519, 693–

712.

Staley, K., Soldo, B. & Proctor, W. (1995). Ionic mechanisms of neuronal excitation by

inhibitory GABAA receptors. Science 269, 977–981.

Stell, B.M., Brickley, S.G., Tang, C.Y., Farrant, M. & Mody, I. (2003). Neuroactive steroids

reduce neuronal excitability by selectively enhancing tonic inhibition mediated by

δ subunit-containing GABAA receptors. Proceedings of the National Academy of

Sciences of the United States of America 100, 14439–14444.

Stern, E.A., Maravall, M. & Svoboda, K. (2001). Rapid development and plasticity of layer

2/3 maps in rat barrel cortex in vivo. Neuron 31, 305–315.

Strömberg, J., Haage, D., Taube, M., Bäckström, T. & Lundgren, P. (2006). Neurosteroid

modulation of allopregnanolone and GABA effect on the GABAA receptor.

Neuroscience 143, 73–81.

Succol, F., Fiumelli, H., Benfenati, F., Cancedda, L. & Barberis, A. (2012). Intracellular

chloride concentration influences the GABAA receptor subunit composition. Nature

Communications 3, 738. doi.org/10.1038/ncomms1744.

Sun, D., Kintner, D.B. & Pond, B.B. (2009). The role of cation-chloride transporters in brain

ischemia. In Physiology and pathology of chloride transporters and channels in the

nervous system, Alvarez-Leefmans, F.J. & Delpire, E. (eds), pp 501-517, Academic

Press, London. Sundgren-Andersson, A.K. & Johansson, S. (1998). Calcium spikes and calcium currents in

neurons from the medial preoptic nucleus of rat. Brain Research 783, 194–209.

Takahashi, T., Kajikawa, Y. & Tsujimoto, T. (1998). G-protein-coupled modulation of

presynaptic calcium currents and transmitter release by a GABAB receptor. Journal of

43

Neuroscience 18, 3138-3146.

Takahashi, T. & Momiyama, A. (1993). Different types of calcium channels mediate central

synaptic transmission. Nature 366, 156–158.

Thompson, S.M. & Gähwiler, B.H. (1989). Activity-dependent disinhibition. II. Effects of

extracellular potassium, furosemide, and membrane potential on ECl- in hippocampal

CA3 neurons. Journal of Neurophysiology 61, 512-523.

Trachtenberg, J.T. & Stryker, M.P. (2001). Rapid anatomical plasticity of horizontal

connections in the developing visual cortex. Journal of Neuroscience 21, 3476–3482.

Tyzio, R., Minlebaev, M., Rheims, S., Ivanov, A., Jorquera, I., Holmes, G. L., … &

Khazipov, R. (2008). Postnatal changes in somatic gamma-aminobutyric acid

signalling in the rat hippocampus. European Journal of Neuroscience 27, 2515–2528.

Tyzio, R., Nardou, R., Ferrari, D.C., Tsintsadze, T., Shahrokhi, A., Eftekhari, S., … & Ben-

Ari, Y. (2014). Oxytocin-Mediated GABA Inhibition During Delivery Attenuates

Autism Pathogenesis in Rodent Offspring. Science 343, 675–679.

Uvarov, P., Ludwig, A., Markkanen, M., Pruunsild, P., Kaila, K., Delpire, E., … &

Airaksinen, M. S. (2007). A novel N-terminal isoform of the neuron-specific K-Cl

cotransporter KCC2. Journal of Biological Chemistry 282, 30570–30576.

Valdivieso, Á.G., Clauzure, M., Massip-Copiz, M. & Santa-Coloma, T.A. (2016). The

chloride anion acts as a second messenger in mammalian cells - modifying the

expression of specific genes. Cellular Physiology and Biochemistry 38, 49–64.

Viitanen, T., Ruusuvuori, E., Kaila, K. & Voipio, J. (2010). The K+-Cl– cotransporter KCC2

promotes GABAergic excitation in the mature rat hippocampus. Journal of Physiology

588, 1527–1540.

Vogt, K. (2015). Diversity in GABAergic signaling. Advances in Pharmacology 73, 203–222.

Voipio, J., Boron, W.F., Jones, S.W., Hopfer, U., Payne, J.A. & Kaila, K. (2014). Comment

on “Local impermeant anions establish the neuronal chloride concentration”. Science

345, 1130.

Vorobjev, V.S. (1991). Vibrodissociation of sliced mammalian nervous tissue. Journal of

Neuroscience Methods 38, 145–150.

Wagner, D.A. (2005). Kinetics and spontaneous open probability conferred by the ε subunit

of the GABAA Receptor. Journal of Neuroscience 25, 10462–10468.

Wagner, S., Sagiv, N. & Yarom, Y. (2001). GABA-induced current and circadian regulation

of chloride in neurones of the rat suprachiasmatic nucleus. Journal of Physiology 537,

853–869.

Watanabe, M., Wake, H., Moorhouse, A.J. & Nabekura, J. (2009). Clustering of neuronal K+-

Cl– cotransporters in lipid rafts by tyrosine phosphorylation. Journal of Biological

Chemistry 284, 27980–27988.

Wilczynski, G.M., Konopacki, F.A, Wilczek, E., Lasiecka, Z., Gorlewicz, A., Michaluk, P.,

… & Kaczmarek, L. (2008). Important role of matrix metalloproteinase 9 in

epileptogenesis. Journal of Cell Biology 180, 1021–1035.

Wlodarczyk, A.I., Sylantyev, S., Herd, M.B., Kersanté, F., Lambert, J.J., Rusakov, D.A., …

& Walker, M.C. (2013). GABA-independent GABAA receptor openings maintain tonic

currents. Journal of Neuroscience 33, 3905–3914.

Wu, H., Che, X., Tang, J., Ma, F., Pan, K., Zhao, M., … & Hong, Y. (2015). The K+–Cl−

cotransporter KCC2 and chloride homeostasis: Potential therapeutic target in acute

central nervous system injury. Molecular Neurobiology 53, 2141-2151.