| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1Department of Anesthesiology, Vanderbilt University Medical Center, Nashville, Tennessee; and 2Laboratory for Integrative Neuroscience, National Institute on Alcohol Abuse and Alcoholism, Rockville, Maryland
Submitted 17 June 2004; accepted in final form 3 October 2004
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
The Cl– concentration could be maintained by secondary active transporters that are mechanisms that use the energy of ion gradients set by other enzymes such as carbonic anhydrase for the generation of a HCO3– gradient and the Na+/K+-ATPase for the generation of Na+ and K+ gradients. Some neurons, most notably adult sensory neurons that exhibit depolarizing GABA responses, have an internal Cl– concentration much higher than electrochemical potential equilibrium. These high Cl– levels are maintained by a strong Na-K-2Cl co-transport activity (Alvarez-Leefmans et al. 1988
; Rohrbough and Spitzer 1996; Sung 2000). In contrast to sensory neurons that keep a high Cl– concentration throughout adulthood, central neurons show high Cl– levels only during development [from E15 to E17 in the ventricular zone (LoTurco et al. 1995; Owens et al. 1996) to postnatal days P1–P7 (Ehrlich et al. 1999; Owens et al. 1996)]. These high Cl– levels are consistent with the depolarizing GABA responses that were measured in immature neurons (Ben-Ari et al. 1989; Janigro and Schwartzkroin 1988; Luhmann and Prince 1991; Mueller et al. 1983; Muller 1989). LoTurco et al. (1995) showed that GABA depolarization in embryonic ventricular zone neurons was related to a furosemide-sensitive Cl– transport process. This process is consistent with an inward Na-K-2Cl co-transporter. Expression of NKCC1, a widely expressed Na-K-2Cl co-transporter, has been shown in immature neurons (Plotkin et al. 1997a) and shown to decrease with neuronal maturation (Plotkin et al. 1997b). Further support for the role of NKCC1 in Cl– accumulation in immature neurons has been provided in recent functional studies (Schomberg et al. 2003).
After maturation, CNS neurons exhibit a Cl– concentration that is lower than electrochemical potential equilibrium, due to an active Cl– transport mechanism that is responsible for extruding Cl– from the neuron. This transporter was shown to be K+-dependent and furosemide-sensitive, consistent with a K-Cl co-transporter (Misgeld et al. 1986; Thompson et al. 1988). In 1996, a neuronal-specific K-Cl co-transporter, KCC2, was cloned (Payne et al. 1996). The expression pattern of KCC2 during development (Clayton et al. 1998; Lu et al. 1999; Rivera et al. 1999) and the phenotype of the KCC2 knockout (Hubner et al. 2001) and knockdown mice (Woo et al. 2002) are all consistent with the important role the co-transporter plays in supporting GABA and glycine hyperpolarizing responses. With the exception of one study that used anti-sense KCC2 oligonucleotide to decrease KCC2 expression and affect EGABA (Rivera et al. 1999), demonstration of the role of KCC2 as a major Cl– regulator in neurons comes mainly from the use of furosemide, an inhibitor of the co-transporter. However, as the loop diuretic lacks specificity, there is a serious need for additional evidence of the role of KCC2 in Cl– regulation in mature central neurons.
In this study, we sought to examine the role of KCC2 in controlling and regulating intracellular Cl– by comparing the reversal potentials of GABAA receptor–mediated Cl– currents in cortical neurons cultured from wild-type and KCC2–/– mice. We show that the normal developmental decrease in [Cl–]i in neurons is absent in mice lacking KCC2 and also show that active Cl– regulation is severely impaired in KCC2–/– neurons, strongly supporting the idea that that KCC2 is critical for Cl– homeostasis in mature cortical neurons.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Due to the early postnatal lethality of the KCC2–/– knockdown animals (Woo et al. 2002), wild-type, heterozygote, and homozygote mice were generated from heterozygote KCC2+/– matings. Genotyping was performed by clipping 1 mm of the tail of newborn (P0–P1) pups. The tail clip was treated with a 200-µl solution containing 25 mM NaOH and 0.2 mM EDTA, pH 
12, for 20 min at 95°C. The sample was neutralized by the addition of a 200-µl solution containing 40 mM Tris-HCl, pH 5. After mixing, the digested tail sample was centrifuged for 6 min at 14,000 rpm, and 200 µl of supernatant was collected for genotyping. Separate PCRs were performed on 1–2 µl tail DNA to amplify fragments specific to the KCC2 control and KCC2 mutant genes. To amplify the control gene, we used forward 5'AGCGTGTGTCCGTGTGCGAGTG-3' and reverse 5'-TTGTTGAGCATGGTGGC TGCGC-3' oligonucleotide primers. To amplify the mutant gene, we used the same forward primer and reverse 5'-CCAGAGGCCACTTGTGTAGCGC-3' primer. Both PCRs generated 200-bp fragments.
Cortical neuronal cultures
Mouse cortical neurons were grown in dissociated cultures according to the method of Huettner and Baughman (1986) with modifications. Briefly, cortices from wild-type KCC2+/+, heterozygotes KCC2+/–, and homozygotes KCC2–/– mouse brains were isolated at postnatal days 0 or 1, and the meninges were removed. Cortices were cut into small pieces and incubated in a 5-ml solution containing 100 U papain, 1 mM L-cysteine, 0.5 mM EDTA, and 500 U DNase1 in Earle’s balanced buffered solution (EBSS) for 60–90 min at 37°C. Following digestion, the tissue was rinsed briefly in EBSS containing 1 mg/ml trypsin inhibitor, 100 U/ml DNase1, and 1 mg/ml bovine serum albumin (BSA) and triturated gently in MEM without serum. Cortical neurons were further purified by centrifugation for 6 min at 70g (room temperature) through a density gradient consisting of 10 mg/ml trypsin and 10 mg/ml BSA in 5 ml EBSS. The cell pellet was resuspended in MEM without serum, counted, and plated at a density of 15,000–40,000 cells on glial feeder layers consisting of growth-arrested rat astrocytes grown on 0.1 mg/ml poly-D-lysine and 33 µg/ml laminin in 35-mm glass bottom dishes. After 3 h, 1.5 ml conditioned growth medium was added to the dishes. The conditioned growth medium consisted of MEM supplemented with 20 mM glucose, 0.5 mM glutamine, 5% fetal bovine serum, 50 U/ml penicillin G, and 50 µg/ml streptomycin and was conditioned by cortical astrocytes as previously described (Baughman et al. 1991). The cultures were kept 
21 days in 5% CO2, 37°C, and fed once a week with 1.5 ml conditioned growth medium.
Immunostaining
Cells were washed with PBS and fixed for 30 min in 2% paraformaldehyde. After fixation, the cell membranes were permeabilized with 0.075% saponin in PBS for 10 min at room temperature (RT), followed by blocking with 0.2% BSA/saponin/PBS for 30 min at RT. Rabbit polyclonal anti-KCC2 antibody (Lu et al. 1999) was diluted 1:200 in BSA/saponin/PBS and incubated with the cells for 1 h at RT. Following several washes in BSA/saponin/PBS, the cells were incubated with goat anti-rabbit immunoglobulin G (Jackson Immunoresearch, West Grove, PA) at a dilution of 1:800 for 1 h at RT and washed several times with BSA/saponin/BSA. For double-staining experiments, cells were incubated successively with anti-KCC2 antibody for 1 h, followed by cy3-conjugated secondary antibody and monoclonal anti-microtubule associated protein-2 (MAP2; clone AP20, Roche, Indianapolis, IN), 1 h at RT, followed by FITC-conjugated anti-mouse immunoglobulin G. Fluorescence signal was visualized using a Zeiss Axiovert S100 microscope equipped with a Photometrics Coolsnap. CCD camera (Roper Scientific, Tucson, AZ) connected to a G4 Apple computer.
Amino acid analysis
The concentration of GABA in the culture medium was assayed by HPLC. To obtain the fluorescent derivatives, 10-µl samples were added to 70 µl of borate buffer and 20 µl 6-aminoquinol-N-hydroxysuccinimidyl carbamate solution (both from AccQ-Tag Chemistry Package kit, WAT052875, Waters, Milford, MA). After heating the mixture for 10 min at 37°C, 10 µl of labeled samples was injected into the HPLC system consisting of a Waters 712 autosampler, two 510 HPLC pumps, a column heater (37°C), and a Waters 474 scanning fluorescence detector. Separation of the amino acids was accomplished using a Waters amino acid column and supplied buffers (buffer A: 19% sodium acetate, 7% phosphoric acid, 2% triethylamine, 72% water; buffer B: 60% acetonitrile), using a specific gradient profile. HPLC control and data acquisition was managed by Millennium 32 software. Using this HPLC solvent system, the following amino acids elute in the following order: cysteine, homocystine, aspartic acid, serine, glutamate, glycine, taurine arginine, threonine, alanine, proline, GABA, cystine, tyrosine, valine, methionine, lysine, isoleucine, leucine, and phenylalanine. Calibration was obtained by running daily calibration curves, consisting of known concentrations of each amino acid (10–100 pmol/µl) to which the internal standard (
-aminobutyric acid, 250 pmol/µl) was added. Peak height of each amino acid was compared with that of the internal standard.
Electrophysiology
Electrophysiological responses of cultured cortical pyramidal neurons were recorded at room temperature using the gramicidin perforated-patch whole cell recording technique with an Axopatch 200 amplifier (Axon Instruments, Foster City, CA) and pCLAMP 8.2 software (Axon Instruments). A 35-mm culture dish of cultured cortical neurons was placed onto an inverted microscope, and an external solution was perfused at a rate of 1 ml/min. The external solution contained 150 mM NaCl, 5 mM KCl, 0.5 mM CaCl2, 1 mM MgCl2, 10 mM glucose, 26 mM sucrose, and 10 mM HEPES (pH 7.4). For the low K+ experiments, the KCl concentration of the external solution was decreased to 1 mM, and the NaCl concentration was increased to 154 mM. Pyramidal neurons were identified by their typical morphology under the microscope. Patch pipette electrodes with resistances of 2–4 M
were made from borosilicate glass capillaries (World Precision Instruments, Sarasota, FL) using a horizontal pipette puller (Sutter Instruments, Novato, CA). The electrodes were first dipped in gramicidin free internal solution consisting of 140 mM KCl, 5 mM EGTA, and 10 mM HEPES, pH adjusted to 7.4 using Tris base, and backfilled with internal solution containing 20–50 µM gramicidin D (Sigma, St. Louis, MO). Gramicidin stock solution (30 µg/ml in DMSO) was kept on ice, and gramicidin-containing internal solution was made fresh every hour to maintain the activity of the ionophore. After the formation of a tight seal, the progress of gramicidin perforation was evaluated by following the series resistance. Recordings started after the series resistance stabilized (15–40 M). Whole cell capacitance (5–30 pF) was compensated before recording. Recordings were low-pass filtered at 2 KHz. Drugs were dissolved in external solution and delivered using a three-barrel SF-77B stepper (Warner Instruments, Hamden, CT) controlled by Pclamp 8.2.
The chloride equilibrium potential was estimated from the reversal potential observed during activation of GABAA receptors (ECl = EGABA). Reversal potentials were recorded using a voltage-ramp protocol consisting of a step to –80 mV for 50 ms, followed by a voltage ramp to –20 mV at a rate of 300 mV/s. At this rate, the current through the membrane capacitance is negligible. The error in EGABA caused by the series resistance was corrected using the following calculation: EGABA = EGABA0 –Rs/Rm x (EGABA0 – Em) (where Rs is the series resistance, Rm is the input resistance, Em is the resting membrane potential, and EGABA0 is the original reversal potential recorded by ramp). Series resistance was compensated by 70–80% in recordings other than voltage ramps.
After each perforated-patch recording, additional negative pressure was applied to the pipette to break the membrane patch under the pipette tip and obtain a whole cell configuration. EGABA was recorded with voltage clamp at different potentials. The value was always measured around 0 mV, indicating absence of voltage drift during the recording. The intracellular chloride ion activity was calculated using the Nernst equation: ECl = RT/F x ln(aCli/aClo), where R is the gas constant, T is the absolute temperature, and F is the Faraday constant (96485.309°C/mol). The extracellular Cl– activity is aClo = [Cl]o, where is the activity coefficient calculated using the extended Debye-Hückel equation. To block voltage-dependent Na+ currents, all experiments were performed in the presence of 0.3 µM TTX (Alomone Labs, Jerusalem, Israel). Furosemide (Sigma) was used at a concentration of 1 mM.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
The postnatal increase in KCC2 expression has been shown in mouse and rat CNS by Northern blot analysis (Lu et al. 1999; Rivera et al. 1999), RNAse protection assay (Clayton et al. 1998; Ganguly et al. 2001), in situ hybridization (Balakrishnan et al. 2003; Clayton et al. 1998), RT-PCR (Balakrishnan et al. 2003), Western blot analysis (Lu et al. 1999; Vu et al. 2000), and immunohistochemistry (Lu et al. 1999; Vu et al. 2000). Using the KCC2-specific antibody developed in our laboratory (Lu et al. 1999; Vu et al. 2000), we observed low KCC2 expression in young immature cortical neurons in culture and high expression levels in older cells (Fig. 1, A and C). Immunostaining of cortical neurons in culture reveals punctate KCC2 expression at the cell plasma membrane of both soma (Fig. 1C) and dendritic spines (Fig. 1C, inset). This expression pattern is absent in KCC2–/– cells (Fig. 1E). Neurons at all stages are easily identified by MAP2 staining (Fig. 1, B, D, and F). The patterns of KCC2 and MAP2 staining are observed in 87% of 284 neurons examined from two wild-type day in culture 14 (DIC14).
|
|
We also measured the resting membrane potential in young and older wild-type and KCC2–/– neurons (Fig. 2B). Although there were no significant differences between the four groups (ANOVA, P > 0.05), we measured a slightly more depolarized potential in older neurons. This small difference, irrespective of the genotype, is consistent with previous reports showing a more depolarized resting membrane potential in cortical neurons of newborn animals (Mienville and Pesold 1999; Zhou and Hablitz 1996). However, this more depolarized potential in young neurons could be due to shunting through the seal resistance, because these cells have a higher input resistance, and no difference in resting membrane potential was measured between young and older neurons when a cell-attached method was used (Tyzio et al. 2003). At resting membrane potential, the size and direction of the GABA current was also determined (Fig. 2, A and C). GABA elicits inward currents in young immature wild-type neurons, and on average, outward currents in older neurons. However, not all older neurons expressed outward GABA currents: a fraction of them still exhibited inward current, despite the fact that 87% of the older neurons showed immunoreactivity with the KCC2 antibody. Consistent with absence of [Cl]i down-regulation in KCC2–/– neurons, GABA-evoked currents remain inward in more mature KCC2–/– neurons (Fig. 2C).
As the neurons differentiate and increase the numbers of synaptic connections in the culture, the concentration of GABA in the culture medium could increase, resulting in increased GABA activity and enhanced Cl– flux across the membrane. However, this is unlikely to occur in the culture system used in this study because glial cells are a major source of ambient GABA, and the glial cells used for the feeder layer in both KCC2+/+ and KCC2–/– cultures were derived from rat. This enhanced tonic GABA activity could account for the observed gradual decrease in aCli toward electrochemical equilibrium. To minimize the accumulation of GABA, we plated the cortical neurons at low density and replaced part of the medium every 5–7 days. To measure the GABA content in the culture, we sampled the culture medium at different time-points and subjected the samples to HPLC. As shown in Fig. 3, the basal level of GABA in the conditioned medium was measured at 6 µM, which increased progressively to 10 µM until replacement of part of the medium every 7 days in both KCC2+/+ or KCC2–/– neuronal cultures. Since the GABA concentration measured in wild-type and KCC2–/– cultures was similar, the absence of aCli decrease in KCC2–/– neurons argues against a role for tonic GABA activity in the gradual decrease that we observed in cultured wild-type control cortical neurons. Taken together, our results show that the Cl– transport mechanism, KCC2, is critical for the decrease in aCl– during postnatal development.
|
Since the Cl– content of mature neurons is significantly below the concentration for passive distribution, energy is required to drive Cl– away from equilibrium. Figures 1 and 2 show that the K+-coupled Cl– transporter, KCC2, is expressed in more mature neurons and is responsible for setting the neuronal Cl– concentration to low values. Whether or not the co-transporter also acutely regulates and maintains intracellular Cl– on challenge is addressed in the next experiments. Using 200-ms ramp protocols, EGABA was measured every 2 min in a DIC14 wild-type neuron. As indicated in Fig. 4A, the measurement procedure does not affect the intracellular Cl– concentration. On addition of furosemide (1 mM), we observed a progressive shift in EGABA toward more positive potentials, indicative of a raise in intracellular Cl–. On removal of the co-transporter inhibitor, the intracellular Cl– returns toward control. The effect of furosemide is significant in older neurons, but not in young neurons, in agreement with the increased expression of KCC2 in the more mature neurons (Fig. 4B). The effect of furosemide in KCC2–/– neurons is examined in later experiments. The driving force for KCC2 results from the difference in the transmembrane ionic gradient of both potassium and chloride. Thus changing the K+ concentration in the recording bath solution directly affects the intracellular Cl– concentration (Kakazu et al. 2000). In the following experiment, after a steady EGABA was observed, the cell perfusion was switched from standard external bath solution containing 5 mM K+ to a bath solution with lower K+ (1 mM, see METHODS). In young wild-type neurons, lowering K+ had little effect on EGABA (Fig. 4C, left; n = 4, 2). In contrast, in older wild-type neurons, exposure to low K+ depleted significantly cell Cl–, resulting in a negative shift of EGABA (Fig. 4C, right; n = 14, 2). This effect of low K+ was counteracted by the addition of 1 mM furosemide (data not shown). The same manipulations were performed in DIC11-13 KCC2–/– neuron cultures, and we observed that the aCli– only slightly decreased when the cells were exposed to low K+ (Fig. 4D, right; n = 8, 1). These data indicate that KCC2 plays a key role in coupling neuronal Cl– with neuronal K+. Other cation-chloride co-transporters, such as KCC1, KCC3, or KCC4, may also participate in the cation-anion coupling, although they may not be active at resting conditions (see DISCUSSION).
|
We have shown that KCC2 plays a critical role in the developmental down-regulation of intracellular Cl–. Because of low internal Cl–, mature neurons are always under the challenge of excessive influx of Cl– ions driven by the large Cl– gradient across the membrane. When GABAA receptors are activated, especially when the neuron is also depolarized, the [Cl]i of small compartments such as dendritic spines can increase significantly through the anion channels associated with GABAA receptors. Different mechanisms, including cation-chloride co-transporters and passive chloride channels, may immediately be activated to fight excessive Cl– influx, and maintain the Cl– equilibrium. To study the role of KCC2 in the acute regulation of [Cl–]i, we used a long 6-s GABA (100 µM) application coupled to a depolarization step (30–40 mV away from the measured resting membrane potential) to induce a large amount of Cl– influx into the recorded neuron. EGABA was monitored using repetitive (every 2 min) brief GABA pulse (80 ms) with voltage-ramp recording. Every experiment started when stable series resistance, and stable EGABAs were achieved.
During each 6-s GABA application under depolarization, the intracellular Cl– significantly increased, as shown by the significant positive shift in EGABA (Fig. 5, B–D). Although the extent to which EGABA changed was different in young and older neurons, the intracellular [Cl–]i returned to its original value within 8–10 min in both groups. The decreases in EGABA during the initial 2-min period were 8.64 ± 2.46 (n = 5) and 1.94 ± 0.52 mV (n = 8) in wild-type DIC14-18 and DIC2-6 neurons, respectively (Fig. 5, D and F). After recovery, a second 6-s GABA application under depolarization was performed, followed in order by an immediate EGABA measurement and the perfusion of 1 mM furosemide. It was only in wild-type neurons older than P10 that furosemide greatly delayed the initial decrease in EGABA (from 8.64 ± 2.46 to 1.76 ± 0.74 mV; n = 5; see Fig. 5, B, D, and F). The loop diuretic failed to affect the initial decrease in EGABA in the young wild-type neurons, in agreement with absence or low KCC2 expression in immature wild-type neurons. Identical experiments were performed in cortical neuronal cultures obtained from KCC2–/– mice. As seen in Fig. 5, C and E, KCC2–/– neurons also recovered their Cl– content after Cl– loading, but the decrease was much smaller (3.12 ± 0.52 mV in DIC11-22 neurons). In these neurons, furosemide did not affect the EGABA decrease (Fig. 5, C, E, and F), consistent with the absence of the co-transporter in these cells. We thus conclude that KCC2 in control neurons actively promotes the rapid extrusion of the anion from the neurons during acute loading of Cl–. Other Cl– mechanisms, probably passive Cl– channels, also take part in the recovery process.
|
Due to the presence of passive chloride permeation pathways, prolonged depolarization induced by persistent activation of GABAergic synapses will tend to shift ECl toward its equilibrium potential. In the next set of experiments, we showed that KCC2 is able to prevent such changes in ECl due to depolarization. Again, cortical neurons obtained from wild-type and homozygote KCC2–/– animals were analyzed. In these experiments, the membrane potential was first clamped to obtain zero current (Erest) and a steady recording of EGABA. The cells were then voltage clamped 30 mV higher than ECl for the entire remainder of the experiment, with the exception of the time during ramping protocols used for EGABA (ECl) measurements. As indicated in Fig. 6, A, C, and D, in older wild-type neurons, the ECl remained steady during depolarization (only a slight increase from –62.4 ± 1.9 to –60.3 ± 1.9 mV). On addition of furosemide, the ECl significantly increased from –60.3 ± 1.9 to –48.6 ± 1.6 mV, indicating a large increase in the intracellular Cl– concentration (Fig. 6D). In contrast, KCC2–/– neurons reacted directly to the depolarization, as indicated by the marked increase in EGABA from 49.2 ± 1.5 to 38.8 ± 1.3 mV (Fig. 6, B, E, and F). As anticipated, in these neurons lacking the co-transporter, furosemide did not affect EGABA. These data show that, in older neurons, KCC2 renders the neuron resistant to changes in the intracellular Cl– concentration during depolarization. Interestingly, in young wild-type neurons, neither depolarization alone nor furosemide significantly affected EGABA (ECl), indicating that only relatively low levels of passive chloride channel pathways are expressed/active in early postnatal development. The absence of a furosemide effect in KCC2–/– neurons not only confirms the absence of the co-transporter in these cells, but also confirms (under our recording conditions) the specificity of the furosemide effect on the co-transporter in wild-type neurons.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
), genetically modified KCC2–/– mice generated in our laboratory survive for 2.5 weeks (Woo et al. 2002). In the brain of these mice, 2–5% KCC2 protein is still detected by Western blot analysis. Northern blot analysis also reveals trace amounts of KCC2 transcript (unpublished data). This leakage is likely due to the targeting of exon 1 and may explain the survival of the mice beyond birth. In their publication, Hubner et al. (2001) showed that EGABA in wild-type E18.5 spinal cord motor neurons is already hyperpolarized (–52 mV), whereas in KCC2–/– neurons, it is more depolarized (–33 mV). Expression of KCC2 is already high in the spinal cord at late embryonic stage (Li et al. 2002; Stein et al. 2004). In our KCC2–/– mouse model, 2–5% KCC2 expression may be enough to slightly shift the EGABA toward the shunting range of GABAA receptor activation, preventing the severe motor deficits that cause neonatal death in the full knockout mice. Of interest is the demonstration here that mature KCC2–/– cortical neurons have lost their ability to regulate [Cl–]i.
Several studies conducted in a variety of GABA-responsive neurons have shown that, from late embryonic days to the second postnatal week, EGABA progressively shifts to more negative potentials (Ben-Ari et al. 1994; Fukuda et al. 1998; Kandler and Friauf 1995; Owens et al. 1996; Wu et al. 1992). This shift is due to a gradual decrease in the neuronal intracellular Cl– content (Fukuda et al. 1998; Owens et al. 1996). Since internal Cl– decreases against an inward driving force, Cl– transport must require energy. The identification of a neuronal-specific K-Cl co-transporter (Payne et al. 1996) and the demonstration of its up-regulation during postnatal neuronal maturation (Clayton et al. 1998; Lu et al. 1999; Rivera et al. 1999), makes the co-transporter a good candidate for this developmental decrease in neuronal Cl–. Indeed, the tight coupling of Cl– and K+ movements through K-Cl co-transport allows for an uphill Cl– transport using the large K+ gradient generated by the Na+/K+ pump. Two additional K-Cl co-transporters could potentially be expressed in cortical neurons. In total, there are four genes encoding K-Cl co-transporters. KCC1 is widely expressed throughout the brain (Payne et al. 1996), where it fulfills housekeeping roles in volume maintenance and regulation. KCC3 expression has been found in large cortical pyramidal cells (Pearson et al. 2001), whereas KCC4 is only expressed in cranial nerves (Karadsheh et al. 2003). Out of the four co-transporters, only KCC2 exhibits substantial basal transport activity (Payne 1997; Song et al. 2002; Strange et al. 2000); the other transporters require swelling activation (Mercado et al. 2000; Mount et al. 1999). In our experimental conditions, there was no indication that an additional K-Cl co-transporter participates in Cl– regulation in older neurons since this regulation was abolished in neurons lacking KCC2.
In agreement with Ganguly et al. (2001), we have shown that isolated neurons reproduce in culture the developmental up-regulation of KCC2 and the concomitant decrease in internal Cl–. An issue that remains unresolved concerns the nature of the mechanism underlying the high Cl– concentration in young immature neurons. Although our data were consistent with many reports showing GABA depolarization in young neurons, we sought to exclude the possibility that the measured high Cl– concentration in the young cultured cortical neurons resulted from a leak from the high Cl–-containing pipette. First, we measured identical GABA reversal potentials in DIC2 neurons when either low Cl– (10 mM) or high Cl– (140 mM) was present in the patch pipette (Fig. 2A). In addition, once the gramicidin perforation was large enough to allow current measurements, we showed that EGABA was very stable despite the high Cl– in the pipette. A progressive leakage of Cl– would have resulted in a progressive shift of EGABA toward more positive potentials. A Cl– activity of 
25 mM in young neurons suggests the participation of an active inward Cl– transport mechanism. Based on expression studies in CNS neurons, we proposed in the past that the Na-K-2Cl co-transporter might participate in Cl– accumulation in immature neurons (Plotkin et al. 1997a, b). However, of seven DIC1-4 neurons, only one responded to 10 µM bumetanide (data not shown). The lack of bumetanide effect is consistent with the absence of effect of decreasing the external K+ concentration (Fig. 4C and DeFazio et al. 2000). These data suggest either that the NKCC1 co-transporter is not expressed or is not active in the young cultured cortical cells or that the Cl– leak in these immature neurons is too small to expose the effect of NKCC1 inhibition. Because bumetanide can directly inhibit GABA receptors (Sung et al. 2000), we did not incubate the neurons in the inhibitor for extended periods of time. The potential participation of NKCC1 in Cl– accumulation can be more directly examined in future studies using an NKCC1 knockout mouse model (Delpire et al. 1999).
In this study, we provide a direct relationship between KCC2 and the developmental decrease in intracellular Cl, because neurons deficient in co-transporter expression fail to undergo this developmental decrease (Fig. 2). The mechanisms underlying the postnatal up-regulation of KCC2 expression are still unknown. We recently showed the presence of a neuronal-restrictive silencing element in proximity of exon 1 (Karadsheh and Delpire 2001). Activation of KCC2 transcription concurs with the decreased expression of neuronal-specific silencing factor after birth (Schoenherr and Anderson 1995). Based on their findings that chronic blockade of GABAA receptors in hippocampal neuronal cultures abolished the KCC2 increase and shift of EGABA, Ganguly et al. (2001) also suggested that the GABA activity promotes the shift of GABA response. However, results from two other groups argue against this effect of chronic GABAA inhibition on KCC2 expression levels (Ludwig et al. 2003; Titz et al. 2003).
In older wild-type neurons (Fig. 2), the internal Cl– concentration is consistent with gradients set by a K-Cl co-transport mechanism: [K+]i x [Cl–]i = [K+]o x [Cl–]o. For known external concentrations of [K+]o at 5 mM and [Cl–]o at 158 mM and estimated [K+]i at 140 mM, the calculated [Cl–]i is 5.6 mM. The [Cl–]i estimated by measurements using gramicidin perforated-patch clamp in DIC21 neurons is 7.3 ± 0.2 mM (n = 4). Thus the agreement between calculated and recorded [Cl–]i indicates that KCC2 is the major mechanism setting [Cl–]i in older neurons. From these data and in agreement with Payne (1997), we can also conclude that, under physiological conditions, KCC2 is working close to its equilibrium. Also consistent with an obligatory coupled K+ and Cl– transport, alteration of extracellular K+ has a significant impact on intracellular Cl– levels. In experiments depicted in Fig. 4, C and D, we showed substantial reduction in Cl– content in older neurons expressing KCC2 on lowering external K+. In young immature neurons, a decrease in [K+]o showed no effect on [Cl–]i. In older homozygous KCC2–/– neurons, the decrease in [K+]o induced a smaller but still significant reduction in [Cl–]i. Whether or not the [Cl–]i reduction is due to redistribution associated with a change in membrane potential or to a K+-coupled mechanism remains to be determined. KCC2 was shown to be highly expressed in the vicinity of excitatory synapses (Gulyas et al. 2001). Thus strong synaptic excitation or pathological conditions leading to increased extracellular K+ will likely result in a significant increase in [Cl–]i. Increased [K+]o and [Cl–]i should both lead to hyperexcitability.
During neuronal inhibition, GABAA receptors are activated, and Cl– is driven into the cell down its electrochemical gradient. During frequent network synaptic activity, intracellular Cl– accumulates due to prolonged GABAA receptor activation, and as a result, EGABA shifts to more positive potentials. During membrane depolarization, Cl– ions can also accumulate through passive chloride conductances (Fig. 5 and Staley et al. 1995). If repeated membrane depolarization is large enough, intracellular Cl– could rise to values where GABA becomes depolarizing. In our experiment shown in Fig. 5, long GABA application combined with depolarization tends to mimic this process. GABAA receptor activation will no longer efficiently inhibit and might even lead to excitation if the accumulated Cl– is not readily removed from the cell. Although other chloride regulating mechanisms such as ClC2 (Staley 1994; Staley et al. 1996) and passive chloride conductance will eventually lower the intracellular Cl–, KCC2 is the major determinant mechanism for the rapid extrusion of excessive Cl– (Fig. 4).
In wild-type neurons expressing high amounts of KCC2, depolarization alone does not affect EGABA, whereas in KCC2–/– neurons, a significant shift of EGABA is observed. These data show that KCC2 is able to regulate intracellular Cl– during membrane depolarization and thus suggest a critical role for KCC2 in efficiently "clamping" intracellular Cl– during neuronal network activities. The activity of KCC2 greatly shortens the recovery time of the neuronal Cl– from excited state to resting state.
In summary, our study provides a comprehensive look at the role of KCC2 in Cl– regulation in isolated CNS neurons. By comparing neurons isolated from genetically modified mice deficient in KCC2 with wild-type neurons, we showed that the co-transporter is critical in lowering intracellular Cl– during postnatal development of cortical principal neurons, a process that occurs progressively during the first 10 days of postnatal life in the forebrain of rodents. We also showed that KCC2 neurons lacking the co-transporter are unable to regulate and maintain their intracellular Cl– during conditions in which the internal Cl– is challenged. In contrast, control neurons use the co-transport of Cl– and K+ to fight Cl– changes completely independently of membrane potential effects (electroneutrality of co-transporter) during GABA inhibition and membrane depolarization. Importantly, we show the absence of a difference in Cl– levels and in acute regulation of Cl– between young neurons from wild-type and KCC2–/– mice, but an increasing difference between the two genotypes during maturation of these neurons. The inability of KCC2–/– cortical primary neurons to regulate Cl– may results in disinhibition and hyperexcitability in the cortical region, but because of the wide distribution of KCC2 within the brain, dysregulation of this co-transporter would likely result in defects involving most brain regions. For instance, disinhibition and hyperexcitability are likely to lead to the generation of seizures (Rivera et al. 2002; Woo et al. 2002). A role for KCC2 in disinhibition and hyperexcitability has also been shown in the spinal cord in a model of chronic pain (Coull et al. 2003).
|
GRANTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
Address for reprint requests and other correspondence: E. Delpire, Dept. of Anesthesiology, Vanderbilt Univ. Medical Center, T-4202 Medical Center North, Nashville, TN 37232-2520 (E-mail: eric.delpire{at}vanderbilt.edu)
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Balakrishnan V, Becker M, Lohrke S, Nothwang HG, Guresir E, and Friauf E. Expression and function of chloride transporters during development of inhibitory neurotransmission in the auditory brainstem. J Neurosci 23: 4134–4145, 2003.
Baughman RW, Huettner JE, Jones KA, and Khan AA. Cell culture of neocortex and basal forebrain from postnatal rats. In: Culturing Nerve Cells, edited by Banker G and Goslin K. Cambridge, MA: MIT Press, 1991, p. 227–249.
Ben-Ari Y, Cherubini E, Corradetti R, and Gaiarsa JL. Giant synaptic potentials in immature rat CA3 hippocampal neurones. J Physiol 416: 303–325, 1989.
Ben-Ari Y, Tseeb V, Raggozzino D, Khazipov R, and Gaiarsa JL. g-Aminobutyric acid (GABA): a fast excitatory transmitter which may regulate the development of hippocampal neurones in early postnatal life. Prog Brain Res 102: 261–273, 1994.[ISI][Medline]
Clayton GH, Owens GC, Wolf JS, and Smith RL. Ontogeny of cation-Cl– co-transporter expression in rat neocortex. Brain Res Dev Brain Res 109: 281–292, 1998.[Medline]
Coull JA, Boudreau D, Bachand K, Prescott SA, Nault F, Sik A, De Koninck P, and De Koninck Y. Trans-synaptic shift in anion gradient in spinal lamina I neurons as a mechanism of neuropathic pain. Nature 424: 938–942, 2003.[CrossRef][Medline]
DeFazio RA, Keros S, Quick MW, and Hablitz JJ. Potassium-coupled chloride co-transport controls intracellular chloride in rat neocortical pyramidal neurons. J Neurosci 20: 8069–8076, 2000.
Delpire E, Lu J, England R, Dull C, and Thorne T. Deafness and imbalance associated with inactivation of the secretory Na-K-2Cl co-transporter. Nat Genet 22: 192–195, 1999.[CrossRef][ISI][Medline]
Ehrlich I, Lohrke S, and Friauf E. Shift from depolarizing to hyperpolarizing glycine action in rat auditory neurones is due to age-dependent Cl– regulation. J Physiol 520: 121–137, 1999.
Fukuda A, Muramatsu K, Okabe A, Shimano Y, Hida H, Fujimoto I, and Nishino H. Changes in intracellular Ca2+ induced by GABAA receptor activation and reduction in Cl- gradient in neonatal rat neocortex. J Neurophysiol 79: 439–446, 1998.
Ganguly K, Schinder AF, Wong ST, and Poo M. GABA itself promotes the developmental switch of neuronal GABAergic responses from excitation to inhibition. Cell 105: 521–532, 2001.[CrossRef][ISI][Medline]
Gulyas AI, Sik A, Payne JA, Kaila K, and Freund TF. The KCl co-transporter, KCC2, is highly expressed in the vicinity of excitatory synapses in the rat hippocampus. Eur J Neurosci 13: 2205–2217, 2001.[CrossRef][ISI][Medline]
Hubner CA, Stein V, Hermans-Borgmeyer I, Meyer T, Ballanyi K, and Jentsch TJ. Disruption of KCC2 reveals an essential role of K-Cl co-transport already in early synaptic inhibition. Neuron 30: 515–524, 2001.[CrossRef][ISI][Medline]
Huettner JE and Baughman RW. Primary culture of identified neurons from the visual cortex of postnatal rats. J Neurosci 6: 3044–3060, 1986.[Abstract]
Janigro D and Schwartzkroin PA. Effects of GABA and baclofen on pyramidal cells in the developing rabbit hippocampus: an in vitro study. Dev Brain Res 41: 171–184, 1988.[CrossRef]
Kakazu Y, Uchida S, Nakagawa T, Akaike N, and Nabekura J. Reversibility and cation selectivity of the K(+)-Cl(–) co-transport in rat central neurons. J Neurophysiol 84: 281–288, 2000.
Kandler K and Friauf E. Development of glycinergic and glutamatergic synaptic transmission in the auditory brainstem of perinatal rats. J Neurosci 15: 6890–6904, 1995.
Karadsheh MF, Byun N, Mount DB, and Delpire E. Localization of the KCC4 potassium-chloride co-transporter in the nervous system. Neuroscience 123: 381–391, 2003.
Karadsheh MF and Delpire E. A neuronal restrictive silencing element is found in the KCC2 gene: molecular basis for KCC2 specific expression in neurons. J Neurophysiol 85: 995–997, 2001.
Li H, Tornberg J, Kaila K, Airaksinen MS, and Rivera C. Patterns of cation-chloride co-transporter expression during embryonic rodent CNS development. Eur J Neurosci 16: 2358–2370, 2002.[CrossRef][ISI][Medline]
LoTurco JJ, Owens DF, Heath MJS, Davis MBE, and Kriegstein AR. GABA and glutamate depolarize cortical progenitor cells and inhibit DNA synthesis. Neuron 15: 1287–1298, 1995.[CrossRef][ISI][Medline]
Lu J, Karadsheh M, and Delpire E. Developmental regulation of the neuronal-specific isoform of K-Cl co-transporter KCC2 in postnatal rat brains. J Neurobiol 39: 558–568, 1999.[CrossRef][ISI][Medline]
Ludwig A, Li H, Saarma M, Kaila K, and Rivera C. Developmental up-regulation of KCC2 in the absence of GABAergic and glutamatergic transmission. Eur J Neurosci 18: 3199–3206, 2003.[CrossRef][ISI][Medline]
Luhmann HJ and Prince DA. Postnatal maturation of the GABAergic system in rat neocortex. J Neurophysiol 65: 247–263, 1991.
Mercado A, Song L, Vazquez N, Mount DB, and Gamba G. Functional comparison of the K+-Cl-co-transporters KCC1 and KCC4. J Biol Chem 275: 30326–30334, 2000.
Mienville JM and Pesold C. Low resting potential and postnatal upregulation of NMDA receptors may cause Cajal-Retzius cell death. J Neurosci 19: 1636–1646, 1999.
Misgeld U, Deisz RA, Dodt HU, and Lux HD. The role of chloride transport in postsynaptic inhibition of hippocampal neurons. Science 232: 1413–1415, 1986.
Mount DB, Mercado A, Song L, Xu J, George JAL, Delpire E, and Gamba G. Cloning and characterization of KCC3 and KCC4, new members of the cation-chloride co-transporter gene family. J Biol Chem 274: 16355–16362, 1999.
Mueller AL, Chesnut RM, and Schwartzkroin PA. Actions of gaba in developing rabbit hippocampus: an in vitro study. Neurosci Lett 39: 193–198, 1983.[CrossRef][ISI][Medline]
Muller D, Oliver M, and Lynch G. Developmental changes in synaptic properties in hippocampus of neonatal rats. Dev Brain Res 49: 105–114, 1989.[Medline]
Owens DF, Boyce LH, Davis MBE, and Kriegstein AR. Excitatory GABA responses in embryonic and neonatal cortical slices demonstrated by gramicidin perforated-patch recordings and calcium imaging. J Neurosci 16: 6414–6423, 1996.
Payne JA. Functional characterization of the neuronal-specific K-Cl co-transporter: implications for [K+]o regulation. Am J Physiol 273: C1516–C1525, 1997.
Payne JA, Stevenson TJ, and Donaldson LF. Molecular characterization of a putative K-Cl co-transporter in rat brain. A neuronal-specific isoform. J Biol Chem 271: 16245–16252, 1996.
Pearson M, Lu J, Mount DB, and Delpire E. Localization of the K-Cl co-transporter, KCC3, in the central and peripheral nervous systems: expression in choroid plexus, large neurons, and white matter tracts. Neuroscience 103: 483–493, 2001.
Plotkin MD, Kaplan MR, Peterson LN, Gullans SR, Hebert SC, and Delpire E. Expression of the Na+-K+-2Cl– co-transporter BSC2 in the nervous system. Am J Physiol 272: C173–C183, 1997a.
Plotkin MD, Snyder EY, Hebert SC, and Delpire E. Expression of the Na-K-2Cl co-transporter is developmentally regulated in postnatal rat brains: a possible mechanism underlying GABA’s excitatory role in immature brain. J Neurobiol 33: 781–795, 1997b.[CrossRef][ISI][Medline]
Rivera C, Li H, Thomas-Crusells J, Lahtinen H, Viitanen T, Nanobashvili A, Kokaia Z, Airaksinen MS, Voipio J, Kaila K, and Saarma M. BDNF-induced TrkB activation down-regulates the K+-Cl- co-transporter KCC2 and impairs neuronal Cl- extrusion. J Cell Biol 159: 747–752, 2002.
Rivera C, Voipio J, Payne JA, Ruusuvuori E, Lahtinen H, Lamsa K, Pirvola U, Saarma M, and Kaila K. The K+/Cl– co-transporter KCC2 renders GABA hyperpolarizing during neuronal maturation. Nature 397: 251–255, 1999.[CrossRef][Medline]
Rohrbough J and Spitzer NC. Regulation of intracellulat Cl– levels by Na+-dependent Cl– co-transport distinguishes depolarizing from hyperpolarizing GABAA receptor-mediated responses in spinal neurons. J Neurosci 16: 82–91, 1996.
), data points were obtained from a total of 180 cells (28 mice). For KCC2–/– cells (
), data points were obtained from a total of 67 cells (7 mice). EGABA (and intracellular Cl– activity) was also determined at DIC2 using low Cl– (10 mM) in the pipette (
). a and b: sample traces showing typical GABA-induced currents during voltage clamp at resting potential. Note that time of GABA application in these experiments is much longer than GABA application in our ramp recordings. c: stable EGABA measured for 
) is very similar to that of wild-type KCC2+/+ cultures (
). Each data point represents the means ± SE GABA concentration in the culture medium (n = 4, KCC2+/+ and n = 2 KCC2–/–).
