EFFECT OF PROTEIN SYNTHESIS INHIBITORS ON THE FUNCTION OF THE NEURONAL K-Cl COTRANSPORTER KCC2

Martin Puskarjov

MASTER’S THESIS

Laboratory of Neurobiology, Major of Physiology, Department of Biosciences,

Faculty of Biological and Environmental Sciences

University of Helsinki

Helsinki 2010

1

Table of Contents List of figures 2 Abbreviations 3 1. Abstract 4 2. Introduction 5

2.1 GABAAR-mediated inhibition 5 2.2 KCC2 renders GABA hyperpolarizing in the mature brain 6 2.3 Regulation of KCC2 function 8

2.3.1 Thermodynamics of KCC2 function 10 2.3.2 Post-translational regulation of KCC2 function 12 2.3.3 Regulation of KCC2 function at the translational level 14

3. Aims of the study 18 4. Materials and methods 19

4.1 Preparation and maintenance of hippocampal slices 19 4.2 Whole-cell somato-dendritic EGABA gradient measurements 20

4.2.1 Electrophysiology 23 4.2.2 Visual identification of apical dendrites and local photolysis of 25 caged GABA

5. Results and discussion 27

5.1 KCC2 function is maintained for hours in the absence of de novo 27 translation 5.2 Somato-dendritic EGABA gradients retain furosemide-sensitivity 32 during inhibited protein synthesis

6. Conclusions 34 7. Acknowledgements 36 8. Bibliography 37

2

List of figures

1. Schematic representation of the life cycle of a plasmalemmal ion transporter

2. The experimental conditions for quantitative assessment of KCC2 function

3. KCC2 function, measured as somato-dendritic ∆EGABA, is suppressed in a manner

that is directly dependent on the duration length of incubation of hippocampal

slices in 100 µM emetine

4. KCC2 function, measured as somato-dendritic ∆EGABA, remains stable for a

minimum of 4 hours under cycloheximide- or emetine-arrested mRNA translation

5. Furosemide sensitivity of somato-dendritic ∆EGABA in the presence of

cycloheximide

3

Abbreviations

ATP adenosine triphosphate BAPTA 1,2-bis(o-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid [Cl-] chloride concentration [Cl-]i intracellular chloride concentration [Cl-]o extracellular chloride concentration CCC cation-chloride cotransporter CGP 55845 (2S)-3-[[(1S)-1-(3,4-Dichlorophenyl)ethyl]amino-2-

hydroxypropyl](phenylmethyl)phosphinic acid CNS central nervous system CNQX 6-cyano-7-nitroquinoxaline-2,3-dione CHX cycloheximide ∆EGABA somato-dendritic difference in reversal potential of GABAA receptor-

mediated responses ∆µion electrochemical potential difference EGABA reversal potential of GABAA receptor-mediated responses Em membrane potential F Faraday’s constant GABA γ-aminobutyric acid GABAAR γ-aminobutyric acid type A receptor GABABR γ-aminobutyric acid type B receptor GHK Goldman-Hodgkin-Katz HEPES 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid IGABA GABAA receptor-mediated current KCC potassium-chloride cotransport KCC2 potassium-chloride cotransporter isoform 2 KCC3 potassium-chloride cotransporter isoform 3 NKCC1 sodium-potassium-chloride cotransporter isoform 1 PKC protein kinase C PP protein phosphatase PP1 type 1 serine/threonine protein phosphatase PTK tyrosine protein kinase R ideal gas constant RBC red blood cell SLC12 solute carrier family 12 T absolute temperature TTX tetrodotoxin UV ultraviolet z valence

4

1. Abstract

The Cl- and HCO3

- electrochemical gradients across the plasma membrane dictate

the electrical consequences of GABAA receptor (GABAAR) function and thereby play a

significant role in neuronal GABA-mediated signalling. In adult pyramidal neurons,

responses to GABA are maintained hyperpolarizing mainly by the action of K-Cl

cotransporter isoform 2 (KCC2). KCC2 acts as a Cl- extrusion mechanism responsible for

setting the intracellular Cl- concentration below the electrochemical equilibrium, a

necessary condition for hyperpolarizing inhibition mediated by GABAARs.

Recent evidence suggests that plasmalemmal KCC2 has a very high rate of

turnover, pointing to a novel role for changes in KCC2 expression in diverse

manifestations of neuronal plasticity. Some studies indicate that rapid down-regulation of

KCC2 may be a general early response involved in various kinds of neuronal trauma.

In this work, whole-cell patch-clamp was used to examine KCC2 function under a

pharmacologically induced arrest of protein synthesis in living hippocampal brain slices

from rat. The stability of KCC2 function was quantitatively assessed on the basis of the

dendritic Cl- extrusion capacity in the presence of protein synthesis inhibitors

cycloheximide and emetine. The parameter used for assessing extrusion capacity was a

somato-dendritic Cl- gradient, which was imposed by a somatic Cl- load that resulted in a

gradient of EGABA (∆EGABA).

The results of this study show that under general protein synthesis inhibitor-

induced arrest of translation, KCC2 function persists unperturbed for at least 4 hours and

hence that the cessation of mRNA translation cannot rapidly induce downregulation of

KCC2-mediated Cl- extrusion. This finding precludes the use of protein synthesis

inhibitors for rapid modulation of KCC2 function. Indirectly, the results presented here

imply that the levels of KCC2 under pathophysiological conditions are primarily

determined by the degradation rate and not by de novo synthesis.

5

2. Introduction

2.1 GABAAR-mediated inhibition

Electrical signaling in excitable cells is generated by dissipating actively

maintained ionic gradients via the opening of neurotransmitter- and voltage-gated ion

channels. Consequently, all brain function boils down to and emerges from the action of

plasmalemmal transport proteins that distribute ions asymmetrically across cell

membranes.

GABA (γ-aminobutyric acid) is the main inhibitory neurotransmitter in the adult

mammalian central nervous system (CNS). It inhibits neuronal firing by activating

ionotropic type A GABA receptors (GABAARs) primarily permeant to chloride (Cl-) and,

to a lesser extent, bicarbonate (HCO3-) ions (Kaila and Voipio 1987; Kaila 1994). Fast

GABAAR-mediated inhibition is defined as a transient decrease in the firing probability

of the target cell (Farrant and Kaila 2007). It is based on the conductance mediated by

GABAARs and on the inward-directed transmembrane electrochemical Cl- gradient that

facilitates net inward Cl- flux upon GABAAR activation. This net Cl- influx manifests as

the major outward current component, in addition to the usually much smaller inward

HCO3- efflux-carried component, of the hyperpolarizing post-synaptic GABAAR-

mediated current (IGABA) that transiently brings the resting membrane potential (Em)

towards the reversal potential of GABAAR-mediated currents (EGABA-A or here simply

EGABA). Provided an intracellular Cl- concentration ([Cl-]i) of ≈ 20 mM or higher, EGABA

is roughly equal to ECl, the reversal potential of the current carried by Cl- (Farrant and

Kaila 2007). The difference between Em and EGABA defines the direction and the driving

force for net Cl- flux and, thus, whether GABAAR activation will lead to

hyperpolarization or to depolarization of the membrane potential.

In the rat hippocampus, fast GABAergic synaptic signaling is depolarizing during

the early postnatal period, becoming hyperpolarizing only by the second postnatal week

(Cherubini et al. 1991). In contrast to the inwardly directed electrochemical gradient of

Cl- in neurons with hyperpolarizing GABAergic responses, an outwardly directed

gradient, that leads to net Cl- efflux upon GABAAR activation, underlies the depolarizing

6

or even excitatory effect of GABA in the early postnatal life and also possibly under

certain post-traumatic conditions of the adult brain (Katchman et al. 1994; van den Pol et

al. 1996; Nabekura et al. 2002).

Transmembrane Cl- concentration gradients, the prerequisite for GABA-mediated

voltage signaling, are a product of active transport. The cation-chloride cotransporters

(CCCs) are membrane proteins that transport Cl- together with the cations sodium (Na+)

and/or potassium (K+) across biological membranes. CCCs do not directly utilize ATP-

derived energy, and are thus secondary active transporters that harness the energy for Cl-

transport from the Na-K ATPase-generated chemical potential differences for Na+ and K+

(Farrant and Kaila 2007). Many members of the CCC protein family, encoded by the

genes Slc12a1-9, are expressed in the CNS (Mercado et al. 2004). However, expression

of one member, the Slc12a5, known as the K-Cl cotransporter isoform 2 (KCC2), is

exclusively restricted to central neurons (Payne et al. 1996; Uvarov et al. 2005 and 2007).

KCC2 functions as a Cl- extrusion mechanism and is responsible for the low [Cl-]i that

underlies the functionally inhibitory effect of GABA (Rivera et al. 1999; Blaesse et al.

2009).

2.2 KCC2 renders GABA hyperpolarizing in the mature brain

Originally studied in the context of volume regulation, K-Cl cotransport (KCC)

has emerged as a pivotal factor that controls the transmembrane Cl- electrochemical

gradient requisite for “classical” hyperpolarizing inhibition mediated by GABAARs and

glycine receptors in excitable cells (Payne et al. 2003). In the vast majority of rat cortical

neurons, KCC2 expression and activity increase during postnatal development (Blaesse et

al. 2009). As net Cl- extrusion is enhanced as a consequence of increased K-Cl

cotransport, a developmental hyperpolarizing shift occurs in the so far depolarizing

GABA- and glycine-elicited post-synaptic potentials.

The unique localization of KCC2 in central neurons (Payne et al. 1996; Uvarov et

al. 2005, 2007) alone is not sufficient to determine KCC2 as the main Cl- extruder

rendering GABA hyperpolarizing in the brain. However, the crucial role of KCC2 in

maintaining hyperpolarizing GABAergic responses has been demonstrated in functional

7

studies manipulating KCC2 expression. As was first demonstrated by Rivera et al.

(1999), reducing the expression levels of KCC2 protein by antisense oligonucleotides

leads to loss of fast hyperpolarizing responses to muscimol, a GABAAR-specific agonist,

in functionally mature hippocampal pyramidal neurons. The significance of KCC2 in

neuronal function was reinforced by Hübner et al. (2001), who demonstrated that

knockout mice deficient in KCC2 die immediately after birth and implied that the

responses to GABA and glycine of their embryonic spinal cord motoneurons are

characteristically excitatory, in contrast to the strictly hyperpolarizing responses of wild

type mice. Recently, a novel KCC2 splice variant was characterized (named KCC2a)

(Uvarov et al. 2007). It differs from the previously characterized KCC2 isoform (now

termed KCC2b) by 40 amino acids in the N-terminus encoded by an alternatively spliced

exon. Mice lacking the KCC2b splice variant but not the KCC2a show neuronal

hyperexcitability and increased susceptibility to epileptic seizures and do not survive

beyond postnatal day 17 (Woo et al. 2002). Mice lacking both KCC2 splice variants die

after birth due to respiratory failure, possibly as a result of malfunction of the respiratory

centre caused by disinhibition (Hübner et al. 2001).

Immature cortical neurons, devoid of active KCC2, characteristically possess

outward transmembrane Cl- electrochemical gradients, maintained via net Cl- uptake by

the Na-K-Cl cotransporter isoform 1 (NKCC1) (Payne et al. 2003). Preventing the

function of NKCC1 with the diuretic bumetanide does not result in the Cl- extrusion

efficacy levels observed in neurons with efficient KCC2-mediated K-Cl cotransport

(Khirug et al. 2010, submitted), which points to that a mere developmental

downregulation of NKCC1 function is unlikely to result in the low [Cl-]i that underlies

the functionally inhibitory effect of GABA in mature hippocampal pyramidal neurons.

Taken together, these studies, which elucidate the importance of KCC2

expression in maintenance of hyperpolarizing GABAergic responses, have implicated

KCC2 as the neuronal Cl- extrusion mechanism responsible for setting the [Cl-]i below

the electrochemical equilibrium, an imperative condition for GABAAR-mediated

hyperpolarizing inhibition (Kaila 1994).

8

2.3 Regulation of KCC2 function

A characteristic feature of GABAAR-mediated signaling is that it can be

qualitatively regulated through short- and long-term changes in the transmembrane anion

gradients, namely those of Cl- and HCO3- (Kaila 1994; Blaesse et al. 2009). This property

of GABAergic signaling has been dubbed “ionic plasticity”. In contrast to short-term

ionic-plasticity, which is based on transient activity-dependent GABAAR-mediated anion

shifts, long-term plasticity arises from changes in expression patterns and kinetic

regulation of the molecular mechanisms responsible for the maintenance of

transmembrane anion gradients (Rivera et al. 2005).

Changes in the capacity of a presynaptic input to influence a postsynaptic output

are thought to be important in learning, memory formation, neural development and

pathological states of the CNS. GABAergic synaptic inhibition plays a critical role in

regulating long-term potentiation (LTP) of glutamatergic synaptic transmission and

circuit output (Gustafsson and Wigström 1988). The synaptic strength or efficacy of an

inhibitory GABAergic synapse is dependent on both the density of GABAARs in the

post-synaptic membrane (Nusser et al. 1998) and on the efficacy of KCC2 function in

generating the driving force for GABAAR-mediated outward currents carried by Cl-. The

polarity of GABAAR-mediated synaptic currents, i.e. whether the activation of

GABAARs results in hyper- or depolarization of the post-synaptic membrane, depends on

the relative membrane levels of functional KCC2 and NKCC1 (Khirug et al. 2008). The

steady-state expression level of membrane-active CCCs may thus play a critical role in

modulating the synaptic input-output relationships. In this context, the high turnover rate

of membrane-bound KCC2, which will be discussed in detail below, points to changes in

KCC2 expression as a possible novel mechanism in the control of neuronal plasticity via

changes in the efficacy of inhibition (Rivera et al. 2004).

9

Figure 1. Schematic representation of the life cycle of a plasmalemmal ion transporter. Ion transporters, like other membrane proteins, are regulated at multiple levels (1-7) in a neuron, for details see text. From Gamba et al. 2009

Ion transporter function is influenced by multiple regulatory mechanisms in the

cell. At the transcriptional/translational levels (1-2, Fig. 1), the spatiotemporal tissue and

cell-type specificity of synthesis of the transporter proteins is determined, followed by the

post-translational level (3-6, Fig. 1), where the activity of the synthesized transporter

protein is controlled by means of reversible events, such as (de)phosphorylation.

Ultimately, the rate of protein degradation (7, Fig 1) delimits the minimum size of the

membrane transporter pool of the cell. Additionally and less obviously, because

cotransport is driven by the electrochemical gradients of the substrate ions, the efficacy of

net transmembrane Cl- cotransport by KCC2 is also, in addition to the abovementioned

regulatory processes (1-7, Fig. 1), influenced by thermodynamic forces.

10

2.3.1 Thermodynamics of KCC2 function

Under physiological conditions, in a cell with an anion permeability, such as the

one provided by GABAARs, but no active chloride transport, Cl- ions will distribute

passively across the cell membrane following the Em of the cell. Since the inside of the

cell is at a negative potential compared to the outside (typical values for a central neuron

range from -50 to -90 mV), an electrical potential difference will drive negatively

charged Cl- ions out of the cell against an inwardly directed Cl- concentration gradient

(more precisely a chemical potential gradient). The sum of the electrical and chemical

potential gradients, the electrochemical potential gradient of Cl- (∆µCl), is the driving

force for Cl- ion flux:

[ ][ ] m

o

iCl FE

ClCl

RT )1(ln −+=∆ −

−

µ , (Eq. 1)

where R is the ideal gas constant, T is the absolute temperature, [Cl-]i/[Cl-]o is the Cl-

transmembrane concentration gradient and F is the Faraday constant. When the force

generated by the chemical potential difference (1st term in Eq. 1) and the force generated

by the electrical potential difference (2nd term in Eq. 1) are equal and opposite, Cl- ions

are subject to no net driving force (∆µCl = 0) and a thermodynamic equilibrium is reached

where no net Cl- flux across the membrane takes place:

[ ][ ] m

o

i EClCl

FRT

=−

−

ln . (Eq. 2)

The above equation is the Nernst equation for Cl- (Nernst 1888), which defines the

transmembrane potential difference arising from a known transmembrane ionic

concentration gradient at electrochemical equilibrium. Under conditions of no “apparent”

active Cl- regulation, the Nernstian equilibrium potential for Cl- (ECl) is equal to Em. It is

important to note here that condition ECl = Em does not rule out active Cl- cotransport.

Because the cellular [Cl-]i is determined by the relative contribution of both the resting

conductance and the active transport machinery, the contribution of an active Cl-

loading/extruding component may be veiled when paralleled with a relatively large Cl-

11

conductance (Kaila et al. 1989; Alvarez-Leefmans and Delpire 2009). From the Nernst

equation, it follows that under equilibrium conditions in total absence of active transport

or in the presence of a relatively weak active transport component, [Cl-]i will change

passively with Em:

[ ] [ ]RT

FEClCl m

oi exp−− = . (Eq. 2.1)

In most cells, including neurons, ECl and [Cl-]i are maintained at a different level

than what is predicted by equations 2 and 2.1, suggesting that in these cells, active Cl-

transport is sufficient enough to maintain Cl- out of thermodynamic equilibrium (Kaila

1994). Neuronal [Cl-]i is maintained at a much lower level than is dictated by passive

distribution via the action of KCC2, which utilizes the outwardly directed Na-K ATPase-

maintained electrochemical potential gradient of K+ to extrude Cl- in a 1:1 stoichiometry.

The driving force (∆µK,Cl) for KCC2-mediated cotransport is determined by the sum of

the electrochemical potential differences of K+ and Cl- (∆µK + ∆µCl):

[ ][ ]

[ ][ ] mCl

o

imK

o

iClK FEz

ClCl

RTFEzKK

RT +++=∆−

−

+

+

lnln,µ (Eq. 3)

Net extrusion of Cl- will take place until the transporter reaches thermodynamic

equilibrium, a situation where ion fluxes mediated by KCC2 do take place, but the net

effluxes and influxes of both K+ and Cl- are equal (Blaesse et al. 2009). KCC2-mediated

transport may proceed in both directions as determined by the electrochemical driving

forces of K+ and Cl- across the membrane. Under physiological ionic conditions transport

is outward and proceeds towards equilibrium, where the net free energy change equals

zero, which for the 1:1 substrate stoichiometry utilized by KCC2 corresponds to EK = ECl

(Payne et al. 2003). Assuming that the system is in thermodynamic equilibrium, i.e.

∆µK,Cl = 0, a simplification of equation 3 yields a definition of equilibrium for K-Cl

cotransport:

[ ] [ ] [ ] [ ]iioo KClKCl +−+− = . (Eq. 3.1)

12

The above equation demonstrates that KCC2 mediated cotransport is driven by the

chemical potential differences of the substrate ions, K+ and Cl-. It also shows that an

increase in [Cl-]i enhances net Cl- extrusion by KCC2 (Payne 1997). The presence of a

strong thermodynamic driving force of substrate ions does not guarantee the activation of

an ion transporter (Blaesse et al. 2009): no transporter may be present in the membrane at

all or it may reside in a state that is incapable of ion transport. To better illustrate this

state, an analogy with a closed ion channel is apt here: despite a large driving force that

may exist across an ion channel, no charge movement across the membrane will take

place if the channel resides in a closed state impermeable to ions. Similarly, a

cotransporter will not move ions across the membrane if it resides in a functionally

inactive state, despite the presence of large driving forces favoring net cotransport. Thus,

although driving force dynamics play a role in regulation of KCC2-mediated cotransport,

the number of ions transported per unit time is not only dependent on the digression of K+

and Cl- transmembrane gradients from thermodynamic equilibrium, but first of all on the

degree of kinetic activation of the transporter, e.g. via phosphorylational mechanisms

(Kaila 1994; Payne et al. 2003; Blaesse et al. 2009).

2.3.2 Post-translational regulation of KCC2 function

Important regulatory mechanisms of KCC2 function at the post-translational level

comprise trafficking of the transporter molecule to the plasma membrane (4, Fig. 1),

insertion of it into the membrane (4, Fig. 1), modulation of the intrinsic ion transport

turnover rates of membrane-bound transporter molecules (5, Fig. 1) and the retrieval of

the transporter from the membrane for recycling or degradation (6 and 7, respectively,

Fig. 1) (Lee et al. 2007; Blaesse et al. 2009; Gamba et al. 2009).

The neuron-specific KCC isoform was first cloned by Payne and co-workers

(Payne et al. 1996), mapping possible targets for kinetic modulation of the cotransporter.

Several protein kinase C (PKC) consensus phosphorylation sites were found to be located

within the hydrophilic intracellular domains, as well as one consensus tyrosine protein

kinase (PTK) phosphorylation site located in the large C-terminus, suggesting

kinase/phosphatase-mediated regulation of KCC2 activity. It is of interest to note here

13

that the abundance of KCC2 protein in a neuron does not necessarily stipulate functional

activity of the cotransporter. For example, the immature neurons in the auditory

brainstem express KCC2 at a high level, yet the transporter is initially functionally

inactive (Blaesse et al. 2006). In addition, inactive KCC2 can be activated rapidly using

the broad-spectrum kinase inhibitor staurosporine in primary cortical cultures during the

first days in vitro, although not in slice preparations (Khirug et al. 2005). A study by

Kelsch et al. (2001) demonstrated that KCC2 activity in cultured hippocampal neurons

required PTK activity, with a notably rapid time-course of KCC2 functional deactivation

within minutes of tyrosine kinase inhibitor application. Furthermore, the co-application of

c-Src, a member of the Src family of kinases, with insulin or insulin-like growth factor

resulted in reactivation of KCC2 in 10 minutes, suggesting the involvement of Src kinase

cascades in fast regulation of functional KCC2.

It is worth noting here that most of the published data regarding kinetic and

allosteric regulation of KCC2 do not explicitly indicate whether changes in the efficacy

of KCC2-mediated cotransport are due to changes in the number of plasmalemmal

transporters or due to the changes in their inherent transport rates, or both. Nevertheless,

a recent paper by Lee et al. (2007) demonstrated that KCC2 membrane surface and

functional expression can be both affected by PKC phosphorylation. Activation of PKC

induced a rapid decrease in the rate of internalization of the cotransporter from the

membrane, increasing cell surface expression levels of KCC2 by 200 % in a mere 10

minutes, as measured using a biotinylation assay. Concurrently, the activity of KCC2,

measured by a 86Rb+ flux assay, was comparably enhanced (Lee et al. 2007). These

findings support the idea of rigorous functional regulation of KCC2 at the post-

translational level via phosphorylation-controlled membrane trafficking.

14

2.3.3 Regulation of KCC2 function at the translational level

The overall turnover of KCC2 protein in the plasma membrane is apparently very

short, measured in terms of tens of minutes (Rivera et al. 2004; Ouardouz and Sastry

2005; Lee et al. 2007). The short membrane half-life of the cotransporter has been

proposed to arise as a result of rapid genomically regulated turnover of KCC2 protein that

is based on continuous kinetic modulation of KCC2 mRNA*, with protein levels of

KCC2 quickly following its mRNA changes (Rivera et al. 2004; Uvarov et al. 2006;

Yang et al. 2009).

If levels of membrane-located KCC2 were to faithfully follow mRNA levels, the

existing total KCC2 protein pool of the cell must be constantly subject to extremely

efficient protein degradation. This suggests that under physiological conditions,

maintenance of the plasmalemmal KCC2 is a balanced steady-state of de novo synthesis

and degradation. Consequently, the experimental prevention of mRNA translation to

KCC2 protein should lead to downregulation of KCC2 function with a time constant

similar to that of the total KCC2 protein pool. Downregulation of KCC2 protein and

function as a result of specific inhibition of KCC2 protein expression, using antisense

oligonucleotides, has been demonstrated (Rivera et al. 1999). In this study, KCC2 total

protein level was shown to dramatically decrease by the 8th hour after specific inhibition

of synthesis, but unfortunately preventing mRNA translation using antisense

oligonucleotides requires long incubation times for the translation block to reach

significant levels, due to poor uptake into cells and degradation of oligonucleotides inside

the cell (Keller et al. 2000), thus precluding conclusions regarding the early degradation

time course.

Intriguingly, a study by Karnani (2008), which addressed the early (0-4 hours)

effect of protein synthesis inhibition on KCC2 protein levels and function, demonstrated

no change in total KCC2 protein levels yet a moderately-rapid downregulation of KCC2

function, in the presence of the protein synthesis inhibitor emetine, with an estimated

* Very little is known about the transcriptional regulatory mechanisms concerning expression of KCC2. Promisingly, analysis of the KCC2 gene has identified several potential transcription factor binding sites in the KCC2 promoter region, including the transcription factor Egr4, which has been shown to have an effect on KCC2 transcription in cultured neurons (Uvarov et al. 2006).

15

half-life of 2.7 hours. Such inconsistency in stability of total protein levels versus

function, observed in the Karnani study, can be explained e.g. by postulating two

differentially regulated KCC2 pools with distinct half-lives. Namely, a large cytosolic

pool with a moderate-to-long half-life, in the range of several hours, and a second smaller

membrane-associated pool that is not replenished from the larger cytosolic pool, with a

faster half-life. Here, the smaller plasmalemmal functional KCC2 could represent such a

small fraction of the total KCC2 pool that its rapid degradation during arrest of protein

synthesis can not be detected in immunoblot of total KCC2 (Karnani 2008). However, the

second protein synthesis inhibitor used in the above study, anisomysin, did not reproduce

the downregulating effect of emetine on KCC2 function. The discrepant result between

the two protein synthesis inhibitors in the Karnani study was attributed to an unspecific

post-translational activating effect on KCC2 function by anisomysin, based on the fact

that under simultaneous application of both inhibitors, anisomysin prevented the emetine-

induced downregulation of KCC2 function. However, such an approach does not exclude

the possibility that the effect on KCC2 function by emetine is also unrelated to protein

synthesis inhibition, but it rather demonstrates that the unspecific effect by anisomysin is

dominant over the one putatively posed by emetine. The results and methodology of the

2008 Karnani study will be discussed in more detail later on, as this study and the present

one are related in their topic and methodological approach.

A rapid protein turnover of KCC2 (cf. Rivera et al. 2004) suggests that the

efficacy of GABAAR-mediated signaling could be rapidly influenced, under both

physiological and pathophysiological conditions, by factors regulating KCC2 gene

expression. Interestingly, the elevation in neuronal [Cl-]i and depolarizing action of

GABAARs that occur after trauma and epileptiform activity have been proposed to result

as a consequence of post-traumatic reduction in KCC2 expression (Rivera et al. 2004;

Bonislawski et al. 2007; Huberfeld et al. 2007). Dramatic loss of up to 80 % of KCC2

protein has been observed after sustained interictal-like activity induced in hippocampal

slices in the absence of Mg2+ (Thomas-Crusells et al. 2000; Rivera et al. 2004),

glutamate-induced excitotoxicity (Chung and Payne 2001) and in a kindling model of

epilepsy (Rivera et al. 2002).

16

Because the steady state level of any protein in a cell is a ratio of the rate of

synthesis and the rate of degradation, it is difficult to discern whether rapid

downregulation of KCC2 protein levels and function and the consequent depolarizing

shift in GABAergic signaling reflect decreased protein synthesis, increased degradation

or both. It remains largely unresolved whether pathological conditions can influence

KCC2 protein level via modulating its native protein synthesis or degradation rates or

both and thus confer KCC2 a differential half-life under post-traumatic versus

physiological conditions. Since the action of translation inhibitors precludes synthesis of

new protein, they can be used as tools to assess to which extent trauma-induced

downregulation of KCC2 can be attributed to diminished de novo protein synthesis.

In the present study, two protein synthesis inhibitors with a similar mode of action

were used in order to reveal any unspecific effects the compounds might pose on the

system under study. The two compounds in this study, emetine, a proemetic alkaloid,

derived from the ipecac root, and cycloheximide (CHX), an antibiotic of bacterial origin

share both structural similarity as well as analogy in inhibitory mode and site of action on

the mammalian translational machinery. Both emetine and CHX inhibit the aminoacyl-

tRNA transfer reaction, preventing the peptide elongation phase during protein

biosynthesis (Grollman 1966). CHX differs from emetine with respect to reversibility of

its effect on protein synthesis; protein synthesis activity can be restored in CHX-treated

cells by removal of the inhibitor from the surrounding medium (Grollman 1966, 1968). In

contrast, the effect of emetine is irreversible; as a consequence, protein synthesis remains

arrested also after removal of this inhibitor from the surrounding medium (Grollman

1968; Stanton and Sarvey 1984). Control experiments indicating complete blockade of

protein synthesis in acute hippocampal slices have been reported for both emetine

(Bradshaw et al. 2003; Fonseca et al. 2006; Karnani 2008) and CHX (Stanton and Sarvey

1984; Kang and Schuman 1996; F. Ahmad, unpublished) at the concentrations used in the

present study.

In summary, the reported high turnover rate of KCC2 protein under both

physiological and pathophysiological conditions (Rivera et al. 2004) should permit the

modulation of neuronal [Cl-]i to take place via changes at the level of KCC2 expression.

Consequently, it should also permit the experimental use of translation inhibiting or

17

degradation enhancing drugs, as tools for rapid modulation of KCC2 function, alongside

drugs modulating protein phosphorylation states already in use for rapid modulation of

KCC2 function at the post-translational regulatory level (e.g. Kelsch et al. 2001; Khirug

et al. 2005; Lee et al. 2007).

18

3. Aims of the study

The K-Cl cotransporter KCC2 is a key molecule in neuronal chloride homeostasis

and has a major impact on GABAergic signaling. KCC2 is modulated via short-term and

long-term changes at the level of transcription, translation, post-translational

modifications, and trafficking. Recent work suggests that a part of total KCC2 protein has

a relatively fast turnover under controlled i.e. non-pathological conditions, with estimated

half-lives ranging from ~15 min up to 2.7 hrs (Rivera et al. 2004, Karnani 2008).

Validation of these estimates could have implications for mechanisms of neuronal

development, plasticity and disease that can be rapidly regulated via genomic

modulations of KCC2 function.

The aims of this Master’s thesis were as follows:

By experimentally preventing neuronal mRNA translation under controlled conditions

1. To investigate the immediate (0-4 hours) degradation time-course of the

functional KCC2 in CA1 pyramidal neurons.

2. To examine the applicability of protein synthesis inhibitors as a method for

modulation of KCC2 function.

3. To compare the effects of different protein synthesis inhibitors on KCC2 function.

19

4. Materials and methods

4.1 Preparation and maintenance of hippocampal slices

The hippocampal formation of the rat (Rattus norvegicus) is anatomically and

physiologically one of the most well-defined brain preparations, serving as a convenient

model for cerebral structure and function. In the present study, the rat hippocampus was

chosen as the experimental preparation because of its well characterized age-related

expression pattern of functional KCC2. In the newborn rat pup, KCC2 expression is very

low, but increases and plateaus by the end of the second postnatal week (Rivera et al.

1999). This transcriptional upregulation is paralleled by a concurrent increase in KCC2

function, which also hits a plateau by postnatal day 14-15 (Khirug et al. 2005). Thus, the

stable baseline activity of KCC2 under normal conditions in the hippocampi of rats aged

15 days and older provides a good model for assessing experimentally-induced changes

in function of KCC2.

Briefly, acute 400 µm coronal hippocampal slices were prepared from postnatal

day 15 and 16 male Wistar rats using methods approved by the University of Helsinki

Animal Care and Use Committee. Prior to decapitation, anesthesia was induced using

halothane (Sigma, Germany). Brains were quickly removed and immersed into an ice-

cold (~3 °C) sectioning solution, containing (in mM): 87 NaCl, 2.5 KCl, 0.5 CaCl2, 25

NaHCO3, 1.25 NaH2PO4, 7 MgSO4, 75 sucrose and 25 D-glucose, equilibrated with

carbogen (95% O2 and 5% CO2), yielding a pH of 7.4 at 32 °C. Tissue containing the

intact hippocampi was dissected and fixed with cyanoacrylate glue (Henkel, Germany)

onto the stage of a Vibratome 3000 Plus (Vibratome, USA) for slicing while immersed in

cooled (~3 °C) sectioning solution. Slices were used within 4 h, after a 1 h recovery

period at 36 °C in recovery solution containing (in mM): 124 NaCl, 3 KCl, 2 CaCl2, 25

NaHCO3, 1.1 NaH2PO4, 2 MgSO4, 6 MgCl2 and 10 D-glucose, continuously bubbled

with carbogen. After the recovery period, the slices were maintained at 32 °C in the

carbogen-bubbled recovery solution until they were placed in a submerged-type chamber

(volume ~1 ml) for recording. Slices were anchored with a stainless steel Lycra-threaded

harp (Warner Instruments, USA) and perfused with a standard extracellular solution

20

(flow rate 1 ml/min), which contained (in mM) 124 NaCl, 3 KCl, 2 CaCl2, 25 NaHCO3,

1.1 NaH2PO4, 2 MgSO4, and 10 D-glucose, bubbled with carbogen (pH 7.4 at 32 °C).

The following drugs were added to the standard extracellular solution in which all

electrophysiological measurements were performed: action potentials were blocked with

0.5 µM TTX (Tocris, UK) and AMPA-type receptors with 10 µM CNQX (Sigma,

Germany), GABAB receptors were blocked with 1 µM CGP 55845 (Tocris, UK) and

NKCC1 was blocked with 10 µM bumetanide (Sigma, Germany) (Russell 2000). To

acutely inhibit KCC2 activity, 1 mM furosemide (Sigma, Germany) was used in addition

to the aforementioned drugs (Payne 1997). Slices were exposed to furosemide for at least

30 minutes before measurements were carried out.

In the protein synthesis inhibitor experiments, hippocampal slices were exposed,

immediately after the end of the 1 h recovery period at 36 °C, to 100 µM cycloheximide

(CHX) or emetine dihydrochloride (Sigma, Germany) in the recovery solution. CHX was

present throughout the experimental procedure also in the standard extracellular solution.

In experiments with emetine, slices were incubated in emetine for 30 minutes in the

recovery solution, after which they were transferred and maintained in emetine-free

recovery solution up until the point of transfer to emetine-free standard solution for

electrophysiological recording. In part of the experiments, as indicated, emetine was

maintained in the extracellular solution throughout the experimental procedure.

4.2 Whole-cell somato-dendritic EGABA gradient measurements

Previous work has shown that plasmalemmal ion transporters are able to maintain

not only transmembrane [Cl-] gradients, but also intracellular [Cl-] gradients along the

dendrites and axons of central neurons. These gradients emerge from the differential

distribution of Na-K- and K-coupled Cl- cotransport mechanisms, mediated by NKCC1

and KCC2, respectively, in the axonal, somatic and dendritic “compartments” of central

neurons. NKCC1 and KCC2 are often coexpressed in mature neurons, and spatially

distinct expression patterns can result in steady-state intraneuronal chloride gradients and

compartmentalization of EGABA within an individual neuron (Szabadics et al. 2006;

Khirug et al. 2008; Blaesse et al. 2009).

21

It appears that, for instance, in the mature dentate granule cells, NKCC1 is

responsible for the depolarizing EGABA values observed in the axon initial segment, where

expression of KCC2 is indetectable (Szabadics et al. 2006). This observation effectively

dismisses the idea of a single EGABA value per neuron, and suggests a dynamic range of

input-specific neuronal EGABA values in the same cell, based on the local relative

efficiency of the two Cl- cotransporters (Khirug et al. 2008; Blaesse et al. 2009).

Consequently, measuring the somatic EGABA value bares no information about other

cellular compartments, where Cl- regulation may be drastically different. The steady-state

EGABA gradients can be quite large, e.g. in cortical principal neurons, the axo-somato-

dendritic EGABA gradient reaches a value up to 15–20 mV (Khirug et al 2008). Most

importantly, measuring the steady-state EGABA in a resting neuron can at best verify the

presence of Cl- extrusion, but bares no information on its efficacy. Provided that net Cl-

influx into a cell i.e. Cl- load is small, even an inefficient cellular Cl- extrusion machinery

will be able to maintain a hyperpolarizing EGABA (Blaesse et al. 2009). The unrelialability

of EGABA as an inidicator of neuronal Cl- extrusion capacity was demonstrated in a study

on chronically injured epileptogenic cortex, which showed that a significant decrease in

KCC2 function was not seen as a change in steady-state EGABA (Jin et al. 2005).

In order to quantitatively assess the efficacy of Cl- extrusion, a defined Cl- load

must be imposed on a cell. The capability of the cell to maintain the [Cl-]i then provides a

valid estimate of the efficacy of extrusion (Jarolimek et al. 1999; Rivera et al. 2004;

Khirug et al. 2005).

Plasmalemmal ion transporters are able to maintain intracellular [Cl-]i gradients

also during a somatic Cl- load under whole-cell conditions (Jarolimek et al. 1999; Khirug

et al. 2005 and 2008; Blaesse et al. 2006; Li et al. 2007). In the whole-cell configuration

of the patch clamp technique (Hamill et al. 1981), where, in addition to forming a high

resistance gigaohm (GΩ) seal between the glass micropipette and a patch of cell

membrane using gentle suction, the sealed patch of membrane is ruptured by additional

application of suction and direct access to the inside of the cell is gained. Since the

volume of the pipette is much greater than that of the cell, the cell interior is dialyzed

with the contents of the electrode (Marty and Neher 1983).

22

In the present study, in order to assess KCC2 function quantitatively, the somato-

dendritic EGABA gradient (∆EGABA = EGABA(soma) - EGABA(dendrite)), a measure of KCC2

functionality, was determined. An artificial defined Cl- load imposed via a patch pipette

clamps the somatic [Cl-] to that of the pipette ([Cl-]pip). As a result of net dendritic Cl-

extrusion by KCC2, a declining somato-dendritic Cl- concentration gradient is formed

from the soma along the dendrite (Fig. 2B).

Figure 2. The experimental conditions for quantified assessment of KCC2 function A Schematic

diagram of the experimental setup used for local photolysis of caged GABA, confocal microscopy, and

somato-dendritic EGABA gradient measurements. From Khiroug et al. 2003. B Schematic diagram of the

somatic Cl- loading assay used to quantify KCC2 function. The somato-dendritic EGABA gradient, measure

of KCC2 functionality, was measured. Briefly, an artificial defined Cl- load imposed via a patch pipette

clamps the somatic [Cl-] to that of the pipette ([Cl-]pip). As a result of net dendritic Cl- extrusion by KCC2, a

declining somato-dendritic Cl- concentration gradient is formed from the soma along the dendrite (indicated

in B as lightening of gray from the soma to the distal dendritic parts). Focal photolytic release of caged

GABA was used to determine the local EGABA (and thus the local [Cl-]i as calculated from the GHK voltage

equation) at the soma (EGABA(soma)) and the dendrite (EGABA(dendrite)). If EGABA(dendrite)<EGABA(soma), an effective

23

Cl- extrusion mechanism is present. Modified from Blaesse et al. 2006. C A confocal micrograph of a

fluorescent CA1 pyramidal neuron loaded with Alexa Fluor 488 through the patch pipette. Pyramidal

neurons were visualized in order to identify a distance of 50 µm from the soma along the apical dendrite,

where the EGABA(dendrite) was measured. A spot of UV light was positioned consecutively at the somatic and

the dendritic location for local GABA uncaging in order to evoke GABAAR-mediated currents. The

potential at which these currents reverse in polarity defines EGABA. Circles indicate the approximate size

and locations of the UV spot used for uncaging. In B and C flashes indicate sites of GABA uncaging.

Focal photolytic release of caged GABA (Fig. 2C) was used to determine the local EGABA

(and thus the local [Cl-]i as calculated from the GHK equation) at the soma (EGABA(soma))

and the apical dendrite (EGABA(dendrite)) 50 µm distance away from the soma. If

EGABA(dendrite)<EGABA(soma), an effective furosemide sensitive Cl- extrusion mechanism is

present (Jarolimek et al. 1999). The magnitude of this gradient along a distance from the

load source directly depends on the local dendritic Cl- extrusion capacity of the cell

(Blaesse et al. 2009).

4.2.1 Electrophysiology

Somatic patch-clamp recordings were carried out at 32 °C in the whole-cell

voltage-clamp configuration using an EPC 10 patch-clamp amplifier (HEKA Elektronik,

Germany) on CA1 pyramidal neurons. Patch pipettes were fabricated from borosilicate

glass (Harvard Apparatus, UK), and their resistances ranged from 4.0 to 6.5 MΩ. Seal

resistances were 2-10 GΩ and pipette capacitance was compensated before breaking the

cell membrane (Marty and Neher 1983). The membrane potential values were corrected

offline for the agar salt bridge-measured liquid junction potential of 10 mV (Barry and

Diamond 1970).

To correctly estimate dendritic Cl- extrusion capacity using the somatic loading

technique, the artificial Cl- load chosen should preferably be close to the concentration

that clamps the somatic [Cl-] ([Cl-]soma) to that of the patch pipette ([Cl-]pipette). If the

chosen Cl- load is subsaturating, part of the Cl- extrusion machinery will remain “idle”,

resulting in an underestimation of extrusion capacity. Thus, too small a load will produce

a false negative error concerning the high Cl- extrusion capacity, i.e. the true capacity is

actually larger than what is given by the measured parameter indicating apparent

24

extrusion capacity. The patch pipette intracellular solution contained (in mM): 18 KCl,

111 K-gluconate, 0.5 CaCl2, 2 NaOH, 10 D-glucose, 10 HEPES, and 2 Mg-ATP,

5 BAPTA (pH 7.3 was adjusted with KOH), yielding a [Cl-]pipette of 19 mM and an EGABA

of -48 mV (Khirug et al. 2005). The Cl- concentration of 19 mM in the patch pipette was

calculated on the basis of the GHK voltage equation (Eq. 3) in order to clamp the somatic

EGABA at approximately -48 mV. Control measurements of the somatic EGABA with

19 mM [Cl-] in the pipette resulted in an average of -49 ± 0.53 mV (n=26), which is

consistent with the theoretically expected value of -48 mV, demonstrating successful Cl-

loading of the soma and clamping of [Cl-]soma to [Cl-]pipette under the present conditions.

Similar somatic EGABA values to the ones observed here, have been reported previously

with the present [Cl-]pipette in hippocampal CA1 pyramidal neurons in acute slices and in

hippocampal neuronal cultures (Khirug et al. 2005) as well in neurons of the lateral

superior olive in the brainstem (Blaesse et al. 2006).

CA1 pyramidal neurons were voltage clamped at a holding potential (VH) of

-50 mV. A holding potential deviant from the resting membrane potential but close to the

calculated EGABA was chosen in order to avoid measuring an EGABA influenced by a large

outwardly directed driving force for Cl- (Akaike et al. 1987), minimizing the leak Cl-

conductance as the driving force (DF) for net transmembrane Cl- flux is close to zero

(DF = VH - EGABA). The reversal potential of GABAergic currents (IGABA = 0) was

determined from the current–voltage (I–V) relation obtained by sequentially clamping the

cell membrane at six different holding potentials with a 5 mV increment and a 10 s

interval between subsequent holding levels. The 4 peak IGABA values with the most

voltage-linearity around the zero-current potential where used to plot the I-V relation.

Step voltages lasted 200 ms and ranged from -65 mV to -35 mV and -70 mV to -40 mV

for somatic and dendritic EGABA measurements, respectively. Caged GABA was

photolyzed to evoke GABAAR-mediated currents 50 ms after the start of each voltage

step, as described below.

Each cell was recorded only once and measurements at both sites were carried out

twice for each cell. A single recording lasted approximately 15 min: after going to whole-

cell mode the somato-dendritic Cl- gradient was allowed to stabilize for 5 min followed

by determination of EGABA(soma), then EGABA(dendrite) with a 2 min interval. After 2 min

25

these measurements were repeated for increased accuracy. The order of measurement of

the somatic and dendritic EGABA had no effect on the measured ∆EGABA gradient (data not

shown).

Only cells with a resting membrane potential more negative than -50 mV were

included in the analysis. It was observed that a low input/series resistance ratio correlates

with drifts in the measured ∆EGABA gradients (data not shown), probably due to poor

somatic Cl- loading of larger volume cells, resulting from clogging of the pipette tip with

cell membrane debris. Thus, only cells with input resistance minimum 10x the series

resistance, were accepted for analysis.

Statistical analysis and curve fitting were performed using Origin (Microcal

Software, USA) and Sigmaplot 10 (Systat, USA) software. Statistical significance was

assessed using the unpaired Student’s t-test. Differences with p < 0.01 were considered

significant. All averaged data are presented as mean ± standard error of the mean.

4.2.2 Visual identification of apical dendrites and local photolysis of caged GABA

In order to visualize the dendrites and accurately measure the distance from the

soma, a fluorescent indicator (Alexa Fluor 488, 100 µM; Invitrogen, USA) was dialyzed

via the patch pipette, and dendrites were traced using a Radiance 2100 confocal

microscope (Bio-Rad, UK) (Fig. 2C). Only cells with an apical dendrite clearly visible in

the XY-plane of stratum radiatum were recorded from, so as to avoid false estimates of

the standard 50 µm somato-dendritic span. Alexa Fluor 488 fluorescence was excited at

488 nm and recorded using a 520 nm long-pass filter.

After visual identification of the correct location for uncaging, GABA was

photolyzed from α-Carboxy-2-nitrobenzyl (CNB)-caged GABA compound (Invitrogen,

Germany) using local uncaging, as described by Khirug et al. (2005). Briefly, the caged

compound was dissolved in a 12 mM stock in physiological solution containing (in mM):

127 NaCl, 3 KCl, 2 CaCl2, 1.3 MgCl2, 20 HEPES, 10 D-glucose, pH 7.4. Before each

experiment, a freshly thawed aliquot of the above stock was diluted to a final

concentration of 2 mM of the caged GABA in the standard extracellular solution and

delivered to the vicinity of the patch-clamped cell at a flow rate of 1-2 µl/min using an

26

UltraMicroPump II syringe pump (WPI, USA) and a syringe with a 100 µm inner tip

diameter (Fig. 2A). For local photolysis of caged GABA, the 375 nm output of a

continuous emission diode laser (Spectra Physics, USA) was delivered to the slice

through an LUMPlanFl 60x water-immersion objective (Olympus, Japan) (Fig. 2A). An

electronic shutter (AA Opto-Electronic, France) was used to set the duration of the laser

pulse at 10-20 ms (Fig. 2A). The UV laser beam was focused to yield an uncaging spot of

~10 µm in diameter (Khirug et al. 2005) either at the soma or at the apical dendrite at a

distance of 50 µm from the soma (Figs. 2B and 2C).

Control experiments were performed (data not shown; cf. Khirug et al. 2008) to

verify that the laser flash evoked no responses in the absence of the CNB-caged GABA

and that caged GABA, when added to the bath solution, did not induce any membrane

current or changes in the input resistance. Furthermore, changing the intensity of the laser

pulse within the range used in this study had no effect on the measured value of EGABA.

27

5. Results and discussion

5.1 KCC2 function is maintained for hours in the absence of de novo translation

Under steady-state conditions, the rate of protein synthesis equals the rate of

protein degradation. Hence, the working hypothesis of the present study was that slice

exposure to protein synthesis inhibitor will prevent the synthesis of new protein

molecules, revealing the protein degradation rate as decay in KCC2 functionality,

provided that protein degradation is preserved under the block of translation (Li et al.

1998; Fonseca et al. 2006; P. Blaesse and B. Haas, unpublished observations). Thus, the

fall in KCC2 expression should be paralleled by a decrease in the neuronal Cl- extrusion

capacity.

The functional half-life of KCC2 was studied by recording the somato-dendritic

∆EGABA gradient of CA1 pyramidal neurons under arrested mRNA translation induced by

protein synthesis inhibitors CHX and emetine. Given that under the present experimental

conditions NKCC1 has been blocked with bumetanide (Payne et al. 2003; Khirug et al.

2008), the furosemide sensitive part of the somato-dendritic ∆EGABA can be attributed to

Cl- extrusion by KCC2 (Li et al. 2007; Khirug et al. 2010), which leads to a more

negative dendritic EGABA relative to the clamped somatic one.

A recent study with a similar approach to assessing the genomically regulated

half-life of the functional membrane-associated KCC2, showed a marked emetine-

induced downregulation of KCC2 function, with an estimated half-life of 2.7 hours

(Karnani 2008). Surprisingly, this study also showed that the total KCC2 pool does not

decrease in parallel with impaired KCC2 function, leading Karnani to speculate about the

existence of two distinct KCC2 pools, with differential turnover rates, to account for the

observed discrepancy. Interestingly, the second protein synthesis inhibitor used in the

Karnani study, anisomycin, did not confirm the emetine effect, which casts doubts on the

genomic origin of the reported half-life of 2.7 hours for the functional pool. What turned

out to be a crucial question regarding the present investigation was whether the

lipophilicity (octanol/water partition coefficient: 2.58) of emetine and irreversibility of its

binding somehow contributes to the downregulation of KCC2 function in an unspecific

28

manner that is unrelated to the inhibitory action of emetine on the neuronal translational

machinery. In the 2008 study by Karnani, exposure of hippocampal slices to 100 µM

emetine was maintained throughout the entire length of the experimental time-window of

4 hours. Keeping in mind the lipophilicity and irreversibility of emetine (Grollman 1968;

Stanton and Sarvey 1984), such approach makes the possibility for unspecific membrane-

action due to high lipophilicity a plausible source of possible unspecific drug effects by

emetine (cf. Pan and Combs 2003).

The present author set forth to test the hypothesis that the hours-on long emetine

application itself and not arrest of protein synthesis per se, is the primary cause for the

moderately rapid downregulation of KCC2 function reported by Karnani (2008). To this

end, hippocampal slices were incubated in 100 µM emetine either for 30 minutes, after

which they were transferred to emetine-free conditions, or alternatively their exposure to

bath-applied emetine was continued until the end of the experiment i.e. up to 4 hours. A

minimum incubation time period of 30 minutes was chosen based on a number of studies

on the protein synthesis-dependent late-phase of LTP, which show that even an

incubation this brief is sufficient to block late-phase LTP (e.g. Stanton and Sarvey 1984;

Huang and Kandel 2006). ∆EGABA measurements from slices without application of any

translation inhibitors, and thus with presumably a fully functioning proteosynthetic

machinery, were performed at the 1st and the 4th hour experimental hour, to investigate

the stability of KCC2 function under control conditions. Mean ∆EGABA values obtained at

the 1st versus the 4th hour did not differ statistically (7.39 ± 0.38 mV/50 µm and

7.68 ± 0.46 mV/50 µm, respectively; p = 0.668; Fig. 4D), providing a stable baseline,

within the present experimental time window of 4 hours. ∆EGABA measurements from

slices initially exposed to emetine for 30 minutes only showed no change in the level of

KCC2 function after 4 hours, as compared to time-matched control slices (7.86 ± 0.68

mV/50 µm; p = 0.823; Fig. 3D). Slices that were incubated for an additional 30 minutes,

thus altogether for 1 hour and slices that were left in emetine until the very end of the 4th

hour, both demonstrated a progressive and highly significant decrease in KCC2 function

(5.53 ± 0.25 mV/50 µm; p = 0.000778 and 3.27 ± 0.29 mV/50 µm; p = 0.0000292,

respectively, as compared to slices incubated for 30 minutes only; Fig. 3D). These data

clearly show that downregulation of KCC2 function, in 100 µM emetine, is not a result of

29

inhibited protein synthesis, but rather an unspecific tissue accumulation-related effect of

long emetine incubation. Because an emetine exposure of only 30 minutes did not lead to

downregulation of ∆EGABA for the minimum of the experimental time period of 4 hours

used herein, but this incubation time is sufficient to irreversibly block protein synthesis at

the concentration used (Stanton and Sarvey 1984), the above data effectively dismiss the

moderately rapid genomically regulated half-life for the functional membrane-located

KCC2 of 2.7 hours proposed by Karnani (2008).

Figure 3. KCC2 function, measured as somato-dendritic ∆EGABA, is suppressed in a manner that is

directly dependent on the duration length of incubation of hippocampal slices in 100 µM emetine.

A-C Sample EGABA recordings at the soma and at a distance of 50 µm in the apical dendrite. Uncaging flash

indicated by horizontal bars. I-V curves, based on uncaging-induced GABAAR-mediated currents (IGABA) at

different holding potentials (Vh; original traces shown as insets here and in subsequent illustrations), used

to determine the somatic and dendritic EGABA values at the 4th hour after a 30 minute-long emetine

incubation (A), at the 1st (B) and at the 4th (C) hour of continuous emetine incubation. D Bars represent

pooled data for recordings from individual cells during the 4th control hour, 4th hour of a 30 minute-long

emetine incubation and during the 1st and the 4th hour of continuous emetine incubation, as indicated. The

30

mean values of ∆EGABA measurements from the 4th hour from slices that were incubated in emetine for 30

minutes only showed no difference from time-matched control slices (p = 0.449). In contrast, ∆EGABA

measurements from the 4th hour from slices that were incubated in emetine for 1 hour demonstrated a

highly significant decrease in KCC2 function (p = 0.000778) compared to mean ∆EGABA values measured at

the 4th hour from slices originally incubated for only 30 minutes. Mean values from slices that were

incubated in emetine until the very end of the 4th hour showed a dramatic additional decrease in ∆EGABA

(p = 0.000029) compared to time-matched slices subjected to an incubation period of 30 minutes only. The

values for n are shown in brackets.

In order to exclude the possibility of lack of an effect on KCC2 function after 2.5

hours by the short 30 minute emetine incubation due to “wearing-off” of the mRNA

translation arrest, as well as to provide additional support for the proposed origin of the

unspecific emetine-effect, hippocampal slices were incubated for 4 hours in the reversible

protein synthesis inhibitor CHX, after which the level of KCC2 function was assessed. In

strict accordance with the ∆EGABA measured at the 4th hour after the short emetine

incubation, the mean ∆EGABA measured from cells exposed to CHX for 4 hours showed

no statistically significant change in the level of KCC2 function from time-matched

control values (8.25 ± 0.23 mV/50 µm; p = 0.449; Fig. 4D). The stability of KCC2

function under both, the continuously present reversible CHX, and the irreversible

emetine applied for 30 minutes, indicates that the lack of effect on KCC2 function by the

short emetine application is a consequence of the fact that cessation of de novo protein

synthesis has no effect on KCC2 function within the present experimental time window

of 4 hours, rather than that of wearing-off of emetine-induced translation block. In sum,

these experiments show that protein synthesis inhibition alone is not sufficient to rapidly

downregulate KCC2 function.

31

Figure 4. KCC2 function, measured as somato-dendritic ∆∆∆∆EGABA, remains stable for a minimum of 4

hours under cycloheximide- (CHX) or emetine-arrested mRNA translation. A-C Sample EGABA

recordings at the soma and at a distance of 50 µm in the apical dendrite. Uncaging flash indicated by

horizontal bars. I-V curves, based on uncaging-induced GABAAR-mediated currents (IGABA) at different

holding potentials (Vh), used to determine the somatic and dendritic EGABA values during the 1st and the 4th

hour under control conditions (A and B, respectively) and during the 4th hour of arrested protein synthesis

induced by CHX (C) and emetine (cf. Fig. 3A). D Bars represent pooled data for recordings from

individual cells during the 1st or the 4th hour, as is individually indicated. The mean values of ∆EGABA

demonstrated no significant change under control conditions between the 1st and the 4th hour (p = 0.668).

Mean ∆EGABA values measured at the 4th hour under both CHX- and emetine-inhibited protein synthesis

were statistically not different from those observed under time-matched control values (p = 0.486 and

p = 0.449, respectively). ∆EGABA values measured under emetine were obtained from slices subjected to 100

µM emetine incubation of 30 minutes after which slices were maintained in emetine-free conditions. CHX

∆EGABA values were obtained from slices incubated in 100 µM CHX throughout the entire length of the

experiment. The values for n are shown in brackets.

32

5.2 Somato-dendritic EGABA gradients retain furosemide-sensitivity during inhibited protein synthesis

In order to confirm that the somato-dendritic EGABA gradients maintained in the

presence of inhibited protein synthesis originate from and are dependent on KCC2-

mediated K-Cl cotransport, the furosemide-sensitivity of the recorded ∆EGABA under the

present experimental conditions was assessed. Furosemide, a diuretic, blocks both

NKCC1 and KCCs with an equal potency (Payne 1997), and it is often used at 100 µM or

higher to inhibit KCC2 in experiments on neuronal cultures and slices (Blaesse et al.

2009; Viitanen et al. 2010). Here, 1 mM furosemide was bath-applied for at least 30 min

before measuring ∆EGABA. Furosemide application alone induced a highly significant

decrease in ∆EGABA from the control value of 7.43 ± 0.25 mV/50 µm to

3.46 ± 0.47 mV/50 µm (p = 0.0000018; Fig. 5C). Furosemide also successfully reduced

the ∆EGABA of CA1 pyramidal neurons exposed for 4 hours to 100 µM CHX to the

furosemide-insensitive level, statistically not different from that observed under control

furosemide levels (3.93 ± 0.66 mV/50 µm; p = 0.624; Fig 5C). These residual

furosemide-insensitive gradients observed under both control conditions and during the

4th hour of CHX incubation are comparable to ∆EGABA gradients previously reported for

KCC2 knockouts (Li et al. 2007) and immature pyramidal neurons (Khirug et al. 2005),

which lack KCC2 activity (Rivera et al. 1999), suggesting that the ∆EGABA gradients

maintained in the presence of inhibited protein synthesis are indeed a product of KCC2-

mediated Cl- extrusion.

33

Figure 5. Furosemide sensitivity of somato-

dendritic ∆EGABA in the presence of cycloheximide

(CHX). A-B Sample EGABA recordings at the soma

and at a distance of 50 µm in the apical dendrite.

Uncaging flash indicated by horizontal bars. I-V

curves, based on uncaging-induced GABAAR-

mediated currents (IGABA) at different holding

potentials (Vh), used to determine the somatic and

dendritic EGABA values under bath application of

furosemide alone (A) and furosemide after a 4

hour incubation in CHX (B). C Bars represent pooled data for recordings from individual cells. Control

and control + furosemide ∆EGABA values where measured at random time points during the 4 hour period

and pooled. Bath application (>30min) of furosemide (1 mM) alone induced a highly significant

downregulation of mean ∆EGABA from control values (p = 0.0000018). Somato-dendritic EGABA gradients

remained furosemide-sensitive in the presence of CHX (100 µM) (p = 0.0000022, as compared to control

∆EGABA values). The furosemide-sensitivity of ∆EGABA under CHX was not statistically different from

furosemide-treated slices not exposed to CHX (p = 0.624). The values for n are shown in brackets.

34

6. Conclusions

The main finding of this Master’s thesis was that de novo protein synthesis is not

a requirement for maintaining KCC2 function for at least 4 hours at a level observed

under physiological conditions in neurons with mature levels of KCC2 expression and

function (Rivera et al. 1999; Khirug et al. 2005). This finding strongly argues against the

idea of special rapid genomic regulation of the membrane-associated functional KCC2,

and suggests that under physiological conditions the functional KCC2 protein does not

have a shorter protein half-life than the total cytosolic KCC2 protein.

Downregulation of KCC2 protein has been shown to take place shortly following

traumatic brain injury (Chung and Payne 2009) and low levels of KCC2 can persist long-

term post brain trauma (Bonislawski et al. 2007). It is entirely possible that trauma-

induced downregulation of KCC2 reflects enhanced degradation of the cotransporter, for

example by protease activation (Chung and Payne 2009). It is intuitively appealing to

assume that trauma-induced loss of KCC2 protein expression in mature neurons involves

both degradation of KCC2 protein and cessation of KCC2 gene expression. The results of

this study clearly show that the latter plays a much smaller role than the former in

downregulating KCC2 function, since the mere prevention of de novo translation did not

lead to rapid loss of neuronal Cl- extrusion capacity. This finding indirectly highlights the

importance of the involvement of the cellular protein degradation in the regulation of Cl-

transport and consequently GABAergic responses, especially in the post-traumatic brain. One of the more technical aims of this thesis work was to examine the

applicability of protein synthesis inhibitors as a method for rapid modulation of KCC2

function. To the best of the knowledge of the present author, the literature does not

identify any effects of cycloheximide unrelated to protein synthesis inhibition that can be

directly linked to post-translational modulation of K-Cl cotransport, on the other hand,

emetine has recently been identified as an activator of PP1 (type 1 serine/threonine

protein phosphatase) (Boon-Unge et al. 2007), which at least in the red blood cells

(RBCs) is a modulator of K-Cl cotransport (Jennings and Schulz 1991; Kaji and

Tsukitani 1991; Bize et al. 1999; De Franceschi et al. 2006). The early findings that

protein phosphatase inhibitor okadaic acid inhibits KCC activity in RBCs (Jennings and

35

Schulz 1991; Kaji and Tsukitani 1991) support the notion that PPs may play a regulatory

role in KCC function. Elevated activity levels of PP1 (Bize et al. 1999) and its isoform

PP1α (De Franceschi et al. 2006) in the erythrocyte membrane are associated with

elevated levels of K-Cl cotransport in RBCs, most likely via action of Src family kinases

(De Franceschi et al. 2006). Although increased PP1 activity clearly induces an

upregulation of KCC function in RBCs, depending on the system, phosphorylation may

result in increased or decreased activity of the target protein (Wera and Hemmings 1995);

consequently, PP1 activation may impose a differential effect on KCC function in

neurons. Thus, the present author proposes that emetine, at high intracellular

concentrations (>100 µM) arising from unspecific effect during incubations longer than

30 minutes, may result in loss of KCC2 function via post-translational mechanisms.

Consequently long incubations at high concentrations should be avoided in experimental

work utilizing emetine to avoid unspecific effects unrelated to protein synthesis

inhibition.

The observed stability of KCC2 function under arrested protein synthesis by the

two protein synthesis inhibitors used in this study, precludes the use of translation

inhibitors for immediate (<4 hours) prevention of KCC2-mediated Cl- extrusion, and thus

their use as tools for rapid modulation of KCC2 function at the genomic level, alongside

drugs modulating protein phosphorylation states (e.g. Kelsch et al. 2001; Khirug et al.

2005; Lee et al. 2007). The results of this study do not exclude, but rather promote the

importance of cellular protein degradation rates in rapid modulation of KCC2 protein

levels and concomitantly function, e.g. in studies modeling the effects of brain trauma on

the inhibitory system.

36

7. Acknowledgements

Without the dedicated guidance and the intellectual support of my supervisors

Professor Kai Kaila and Dr. Peter Blaesse, and the theoretical and practical lessons in

electrophysiology given to me by Professor Juha Voipio, Dr. Ramil Afzalov, MSc

Stanislav Khirug and Dr. Eva Ruusuvuori, the completion of this Master’s thesis could

not have been possible. I would like to thank all the members of the Laboratory of

Neurobiology, especially MSc Faraz Ahmad, for our discussions on things that first seem

simple but always turn out complicated in the end.

37

8. Bibliography

Akaike,N., Inomata,N. and Tokutomi,N. (1987). Contribution of chloride shifts to the fade of gamma-aminobutyric acid-gated currents in frog dorsal root ganglion cells. J Physiol (Lond) 391, 219-234. Alvarez-Leefmans,F., and Delpire,E. (2009), Thermodynamics and kinetics of chloride transport in neurons: an outline. In: Alvarez-Leefmans,F.J. and Delpire,E. (Eds), Physiology and pathology of chloride transporters and channels in the nervous system. Academic Press, London, 81-108. Barry,P.H. and Diamond,J.M. 1970, Junction potentials, electrode standard potentials, and other problems in interpreting electrical properties of membranes, JMB 3, 93-122. Bize,I., Guvenc,B., Robb,A., Buchbinder,G. and Brugnara,C. (1999). Serine/threonine protein phosphatases and regulation of K-Cl cotransport in human erythrocytes. Am J Physiol 277, C926–C936. Blaesse,P., Airaksinen,M.S., Rivera,C., and 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., Friauf,E., and Nothwang,H.G. (2006). Oligomerization of KCC2 correlates with development of inhibitory neurotransmission. J Neurosci 26, 10407-10419.

Bonislawski,D.P., Schwarzbach,E.P., and Cohen,A.S. (2007). Brain injury impairs dentate gyrus inhibitory efficacy. Neurobiol Dis 25, 163-169. Boon-Unge,K., Yu,Q., Zou,T., Zhou,A., Govitrapong,P. & Zhou,J. (2007). Emetine regulates the alternative splicing of Bcl-x through a protein phosphatase 1-dependent mechanism. Chembiol 14, 1386-1392. Bradshaw,K.D., Emptage,N.J. and Bliss,T.V.P. (2003). A role for dendritic protein synthesis in hippocampal late LTP. Eur J Neurosci 18, 3150-3152. Cherubini,E., Gaiarsa,J.L. and Ben-Ari,Y. (1991). GABA - an excitatory transmitter in early postnatal life. Trends Neurosci 14, 515-519. Chung,C. and Payne,J.A. (2001) Glutamate excitotoxicity causes rapid degradation of the neuronal K-Cl cotransporter (KCC2) via a calcium-activated neutral protease. Soc Neurosci Abstr 27, 710.1. Chung,C.Y. and Payne,J.A. (2009). Modulation of GABAergic signaling: role of neuronal chloride. In: Encyclopedia of Basic Epilepsy Research, (Ed P.A. Schwartzkroin). Elsevier.

38

Franceschi,L.D., Villa-Moruzzi,E., Biondani,A., Siciliano,A., Brugnara,C., Alper,S.L., Lowell,C.A., and Berton,G. (2006). Regulation of K-Cl cotransport by protein phosphatase 1alpha in mouse erythrocytes. Pflűgers Arch 451, 760-768. Farrant,M., and Kaila,K. (2007). The cellular, molecular and ionic basis of GABAA receptor signalling. Prog Brain Res 160, 59-87.

Fonseca,R., Vabulas,R.M., Hartl,F.U., Bonhoeffer,T. and Nagerl,U.V. (2006). A balance of protein synthesis and proteasome-dependent degradation determines the maintenance of LTP. Neuron 52, 239-245. Gamba,G., Garbarini,N. and Delpire,E. (2009). Regulation of cation-chloride cotransporters. In: Alvarez-Leefmans,F.J., Delpire,E. (Eds), Physiology and pathology of chloride transporters and channels in the nervous system. Academic Press, London, 357-381. Grady,M.S. and Lam,A.M. (1995). Management of acute head injury. In: Lam,A.M. (Ed.), Anesthetic Management of Head Injury. McGraw-Hill, NY, 87-100. Grollman,A.P. (1966). Structural basis for inhibition of protein synthesis by emetine and cycloheximide based on an analogy between ipecac alkaloids and glutarimide antibiotics. Proc Nat Acad Sci USA 56, 1867-1874. Grollman,A.P. (1968). Inhibitors of protein biosynthesis. V. Effects of emetine on protein and nucleic acid biosynthesis in HeLa cells. J Biol Chem 243, 4089-4094. Gustafsson,B. and Wigström,H. (1988). Physiological mechanisms underlying long-term potentiation. Trends Neurosci 11, 156-162. Hamill,O.P., Marty,A., Neher,E., Sakmann,B., and Sigworth,F.J. (1981). Improved patch-clamp techniques for high-resolution current recording from cells and cell-free membrane patches. Pflűgers Arch 39, 85-100. Huang,Y.Y. and Kandel,E.R. (2006). Age-related enhancement of a protein synthesis-dependent late phase of LTP induced by low frequency paired pulse stimulation in hippocampus. Learn Mem 13, 298-306. Huberfeld,G., Wittner,L., Clemenceau,S., Baulac,M., Kaila,K., Miles,R., and Rivera,C. (2007). Perturbed chloride homeostasis and GABAergic signaling in human temporal lobe epilepsy. J Neurosci 27, 9866–9873. Hűbner,C.A., Stein,V., Hermans-Borgmeyer,I., Meyer,T., Ballanyi,K., and Jentsch,T.J. (2001). Disruption of KCC2 reveals an essential role of K-Cl cotransport already in early synaptic inhibition. Neuron 30, 515–524.

39

Jarolimek,W., Lewen,A. and Misgeld,U. (1999). A furosemide-sensitive K-Cl cotransporter counteracts intracellular Cl- accumulation and depletion in cultured rat midbrain neurons. J Neurosci 19, 4695-4704. Jennings,M.L. and Schulz,R.K. (1991). Okadaic acid inhibition of K-Cl cotransport. Evidence that protein dephosphorylation is necessary for activation of transport by either cell swelling or N-ethylmaleimide. J Gen Physiol 97, 799-818. Jin,X., Huguenard,J.R., and Prince,D.A. (2005). Impaired Cl- extrusion in layer V pyramidal neurons of chronically injured epileptogenic neocortex. J Neurophysiol 93, 2117-2126. Kaila,K. (1994). Ionic basis of GABAA receptor channel function in the nervous system. Prog Neurobiol 42, 489-537.

Kaila,K., Pasternack,M., Saarikoski,J. and Voipio,J. (1989). Influence of GABA-gated bicarbonate conductance on membrane potential, current and intracellular chloride in crayfish muscle fibres. Journal of Physiology 416, 161-181. Kaila,K. and Voipio,J. (1987). Post-synaptic fall in intracellular pH induced by GABA-activated bicarbonate conductance. Nature 330, 163-165. Kaji,D.M. and Tsukitani,Y. (1991). Role of protein phosphatase in activation of KCl cotransport in human erythrocytes. AJP - Cell Physiology 260, C176-180. Kang,H. and Schuman,E.M. (1996). A requirement for local protein synthesis in neurotrophin-induced hippocampal synaptic plasticity. Science 273, 1402-1406. Karnani,M. (2008). KCC2 Protein pools, and the role of KCC2 turnover in rapid regulation of neuronal Cl- transport. Master’s thesis, Helsinki University. Katchman,A.N., Vicini,S., and Hershkowitz,N. (1994). Mechanism of early anoxia-induced suppression of the GABAA-mediated inhibitory postsynaptic current. J. Neurophysiol 71, 1128–1138. Keller,H., Schu,B., Rudel,R. and Brinkmeier,H. (2000). Cellular uptake and efficacy of antisense oligonucleotides against RNAs of two Na+ channel isoforms. J Pharmacol Exp Ther 295, 367-372. Kelsch,W., Hormuzdi,S., Straube,E., Lewen,A., Monyer,H. and Misgeld,U. (2001). Insulin-like growth factor 1 and a cytosolic tyrosine kinase activate chloride outward transport during maturation of hippocampal neurons. J Neurosci 21, 8339-8347. Khiroug,L., Giniatullin,R., Klein,R.C., Fayuk,D. and Yakel,J.L. (2003). Functional mapping and Ca2+ regulation of nicotinic acetylcholine receptor channels in rat hippocampal CA1 neurons. J Neurosci 23, 9024-9031.

40

Khirug,S., Ahmad,F., Afzalov,R., Puskarjov,M., Khiroug,L., Kaila,K., and Blaesse,P. (2010). A single seizure episode leads to rapid functional activation of KCC2 in the neonatal rat hippocampus. Under revision.

Khirug,S., Huttu,K., Ludwig,A., Smirnov,S., Voipio,J., Rivera,C., Kaila,K., and Khiroug,L. (2005). Distinct properties of functional KCC2 expression in immature mouse hippocampal neurons in culture and in acute slices. Eur J Neurosci 21, 899-904.

Khirug,S., Yamada,J., Afzalov,R., Voipio,J., Khiroug,L., and Kaila,K. (2008). GABAergic depolarization of the axon initial segment in cortical principal neurons is caused by the Na-K-2Cl cotransporter NKCC1. J Neurosci 28, 4635-4639.

Lee,H.H., Walker,J.A., Williams,J.R., Goodier,R.J., Payne,J.A. and Moss,S.J. (2007). Direct protein kinase C-dependent phosphorylation regulates the cell surface stability and activity of the potassium chloride cotransporter KCC2. J Biol Chem 282, 29777-29784. Li,H., Khirug,S., Cai,C., Ludwig,A., Blaesse,P., Kolikova,J., Afzalov,R., Coleman,S.K., Lauri,S., Airaksinen,M.S., Keinanen,K., Khiroug,L., Saarma,M., Kaila,K., and Rivera,C. (2007). KCC2 interacts with the dendritic cytoskeleton to promote spine development. Neuron 56, 1019-1033.

Li,X., Zhao,X., Fang,Y., Jiang,X., Duong,T., Fan,C., Huang,C-C., and Kain,S.R. (1998). Generation of destabilized green fluorescent protein as a transcription reporter. J Biol Chem 273, 34970–34975. Marty,A. and Neher,E. (1983). Tight-seal whole-cell recording. In: Sakmann,B. and Neher E. (Eds) Single channel recording. Plenum Press, New York, 107-122. Mercado,A., Mount,D.B. and Gamba,G. (2004). Electroneutral cation-chloride cotransporters in the central nervous system. Neurochem Res. 29, 17-25. Nabekura,J., Ueno,T., Okabe,A., Furuta,A., Iwaki,T., Shimizu-Okabe,C., Fukuda,A., and Akaike, N. (2002). Reduction of KCC2 expression and GABAA receptor-mediated excitation after in vivo axonal injury. J Neurosci 22, 4412–4417. Nernst,W. (1888). Zur Kinetic der in Lösung befindlichen Körper: Theorie der Diffusion. Z Phys Chem, 613-637. Nusser,Z. Hajos,N. Somogyi,P. Mody,I. (1998). Increased number of synaptic GABAA receptors underlies potentiation at hippocampal inhibitory synapses. Nature 395, 172–177. Ouardouz,M. and Sastry,BR. (2005) Activity-mediated shift in reversal potential of GABAergic synaptic currents in immature neurons. Brain Res 160, 78-84. Pan,C.J. and Combs,A.B. (2003). Emetine inhibits glycolysis in isolated, perfused rat hearts. Cardiovasc Toxicol 3, 311-318.

41

Payne,J.A. (1997). Functional characterization of the neuronal-specific K-Cl cotransporter - implications for [K+]o regulation. Am J Physiol 42, C1516-C1525. Payne,J.A., Rivera,C., Voipio,J., and Kaila,K. (2003). Cation-chloride co-transporters in neuronal communication, development and trauma. Trends Neurosci 26, 199-206.

Payne,J.A., Stevenson,T.J. and Donaldson,L.F. (1996). Molecular characterization of a putative K-Cl cotransporter in rat brain. A neuronal-specific isoform. J Biol Chem 271, 16245-16252. Rivera,C., Li,H., Thomas-Crusells,J., Lahtinen,H., Viitanen,T., Nanobashvili,A., Kokaia,Z., Airaksinen,M.S., Voipio,J., Kaila,K., and Saarma,M. (2002). BDNF-induced TrkB activation down-regulates the K+-Cl- cotransporter KCC2 and impairs neuronal Cl- extrusion. J Cell Biol 159, 747-752.

Rivera,C., Voipio,J. and Kaila,K. (2005). Two developmental switches in GABAergic signaling: the K-Cl cotransporter KCC2 and carbonic anhydrase CAVII. J Physiol 562, 27-36. Rivera,C., Voipio,J., Payne,J.A., Ruusuvuori,E., Lahtinen,H., Lamsa,K., Pirvola,U., Saarma,M., and 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., Sipilä,S., Payne,J.A., Minichiello,L., Saarma,M., and Kaila,K. (2004). Mechanism of activity-dependent downregulation of the neuron-specific K-Cl cotransporter KCC2. J Neurosci 24, 4683-4691.

Russell,J.M. (2000). Sodium-potassium-chloride cotransport. Physiol Rev 80, 211–276. Stanton,P.K. and Sarvey,J.M. (1984). Blockade of long-term potentiation in rat hippocampal CA1 region by inhibitors of protein synthesis. J Neurosci 4, 3080-3088. Szabadics,J., Varga,C., Molnar,G., Olah,S., Barzo,P., and Tamas,G. (2006). Excitatory effect of GABAergic axo-axonic cells in cortical microcircuits. Science 311, 233–235. Thomas-Crusells,J., Vieira.A., Airaksinen,M.S., Saarma,M., Kaila,K. and Rivera,C. (2000). Downregulation of the neuronal K–Cl cotransporter KCC2 in basal CA1 dendrites following epileptic activity in vitro is a consequence of a fast turn-over rate and slowing down of KCC2 protein transport. Soc Neurosci Abstr 26, 233.236. Uvarov,P., Pruunsild,P., Timmusk,T., and Airaksinen,M.S. (2005). Neuronal K+/Cl- cotransporter (KCC2) transgenes lacking neurone restrictive silencer element recapitulate CNS neurone-specific expression and developmental up-regulation of endogenous KCC2 gene. J Neurochem 95, 1144-1155.

42

Uvarov,P., Ludwig,A., Markkanen,M., Pruunsild,P., Kaila,K., Delpire,E., Timmusk,T., Rivera,C. and Airaksinen,M.S. (2007). A novel N-terminal isoform of the neuron-specific K-Cl cotransporter KCC2. J Biol Chem 282, 30570-30576. van den Pol,A.N., Obrietan,K., and Chen,G. (1996). Excitatory actions of GABA after neuronal trauma. J Neurosci 16, 4283-4292. Viitanen,T., Ruusuvuori,E., Kaila,K. and Voipio,J. (2010). The K+-Cl- cotransporter KCC2 promotes GABAergic excitation in the mature rat hippocampus. J Physiol, in press. Wera,S. and Hemmings,B.A. (1995). Serine/threonine protein phosphatases. Biochem J 311, 17-29. Woo,N.S., Lu,J.M., England,R., McClellan,R., Dufour,S., Mount,D.B., Deutch,A.Y., Lovinger,D.M., and Delpire,E. (2002). Hyperexcitability and epilepsy associated with disruption of the mouse neuronal-specific K-Cl cotransporter gene. Hippocampus 12, 258-268. Yang, B., Tadavarty,R., Xu,J.Y. and Sastry,B.R. (2010). Activity-mediated plasticity of GABA equilibrium potential in rat hippocampal CA1 neurons. Exp Neurol 221, 157-165.