Ammonium influx pathways into astrocytes and ... - Wiley Online Library

0 downloads 0 Views 753KB Size Report
cotransport; NKCC, Na+-dependent KCC; SR101, sulforhodamine 101. ... +, Na+,K+-ATPase is the major influx pathway for NH4. + in ...... and role in signalling.
JOURNAL OF NEUROCHEMISTRY

| 2010 | 115 | 1123–1136

doi: 10.1111/j.1471-4159.2010.07009.x

1

Institut fu¨r Neurobiologie, Heinrich-Heine-Universita¨t Du¨sseldorf, Du¨sseldorf, Germany

Abstract Ammonium (NH4+) is required to maintain pathways involved in shuttling metabolic precursors between astrocytes and neurones. Under hyperammonaemic conditions, increases in the cellular influx of NH4+, and accompanying changes in ion concentrations, may contribute to disruptions in metabolism and neurotransmission. We investigated mechanisms of cellular NH4+ influx in hippocampal slices by measuring acute NH4+/NH3-evoked changes in intracellular pH (pHi) and sodium ([Na+]i). In both astrocytes and neurones, application of 5 mM NH4Cl for 30–45 min decreased pHi by 0.2–0.3 units, consistent with NH4+ influx. In astrocytes, but not neurones, acidifications were accompanied by [Na+]i increases of 25–30 mM. Glial [Na+]i increases were blocked by bumetanide, suggesting that NH4+/NH3 activated Na+-dependent, K+, Cl) cotransport.

Bumetanide also reduced NH4+/NH3-evoked acidifications in astrocytes. Neuronal acidifications were insensitive to bumetanide and inhibition of Cl)-dependent transport and K+ channels, but were prevented by inhibition of Na+,K+-ATPase with ouabain. Furthermore, ouabain reduced astrocyte acidifications. Our results suggest that following rapid elevation of NH4+, Na+,K+-ATPase is the major influx pathway for NH4+ in neurones, whereas Na+,K+-ATPase and Na+-dependent, K+, Cl) cotransport mediate NH4+ transport into astrocytes. The different mechanisms of NH4+ influx in astrocytes and neurones may contribute to the different susceptibility of both cell types to acute hyperammonaemic conditions. Keywords: hepatic encephalopathy, hyperammonaemia, Na+, K+-ATPase, NKCC, pH, sodium. J. Neurochem. (2010) 115, 1123–1136.

Hepatic encephalopathy (HE) is a complex neuropsychiatric syndrome resulting from acute or chronic liver dysfunction (see Weissenborn et al. 2005). Ammonium, in conjunction with unprotonated ammonia (NH4+/NH3), is the leading candidate responsible for the various neurological dysfunctions observed during HE. Following acute liver failure, NH4+/NH3 can reach 5 mM in the CNS. Moreover, concentrations of NH4+/NH3 in the CNS correlate with the symptoms of acute HE, which range from altered sleep patterns and mild confusion in grade I HE, to coma and brain herniation resulting in death in grade IV HE (see Felipo and Butterworth 2002; Ytrebo et al. 2009). Under physiological conditions, NH4+/NH3 concentrations are < 0.1 mM in the cerebrospinal fluid of mammals (Marcaggi and Coles 2001). NH4+/NH3 is essential for recycling neurotransmitters via the glutamate/GABA-glutamine cycle in the brain (Bak et al. 2006). In addition to maintaining the glutamate/GABA-glutamine cycle, the influx of NH4+ into astrocytes and its subsequent use in the conversion of glutamate to glutamine is the sole pathway for the detoxification of elevated levels of NH4+ in the CNS. During hyperammonaemia, a number of cellular processes

are altered including excitatory neurotransmission (Monfort et al. 2002; Ahboucha and Butterworth 2004). Increased extracellular glutamate concentration and the activation of post synaptic glutamate receptors are widely viewed as central to the pathology of HE (see e.g. Rose 2002). In this regard, we recently found that extracellular application of NH4+/NH3 inhibits glutamate uptake into cultured astrocytes within 20–30 min. The inhibitory action of NH4+/NH3 was primarily caused by influx of NH4+ via activation of Na+, K+, Cl) co-transport: the accompanying pHi decreases and [Na+]i increases reduced the inward transmembrane gradients of

Received April 28, 2010; revised manuscript received September 3, 2010; accepted September 3, 2010. Address correspondence and reprint requests to Christine R. Rose, Institut fu¨r Neurobiologie, Heinrich-Heine-Universita¨t Du¨sseldorf, Universita¨tsstraße 1, Geba¨ude 26.02.00, 40225 Du¨sseldorf, Germany. E-mail: [email protected] 1 Present address: Laboratory for Cognition Research and Experimental Epileptology, Department of Epileptology, University of Bonn, Sigmund-Freud-Strasse 25, D-53105 Bonn, Germany. Abbreviations used: HE, hepatic encephalopathy; KCC, K+, Cl) cotransport; NKCC, Na+-dependent KCC; SR101, sulforhodamine 101.

 2010 The Authors Journal of Neurochemistry  2010 International Society for Neurochemistry, J. Neurochem. (2010) 115, 1123–1136

1123

1124 | T. Kelly and C. R. Rose

Na+ and H+, which provide the driving force for glutamate uptake (Kelly et al. 2009). In addition, the influx of NH4+ is a likely first step in affecting intracellular enzyme pathways and disruption of mitochondrial function. For instance, the breakdown of glutamate via the tricarboxylic acid cycle was reduced by NH4+/NH3, likely because of the inhibition of a-ketoglutarate dehydrogenase in both cultured astrocytes and neurones (see Hertz and Kala 2007). Furthermore, following the cellular influx of NH4+, glutamine production and its subsequent breakdown in mitochondria may increase free radical production, resulting in mitochondrial dysfunction, and contribute to the metabolic deficits observed in HE patients (see Jayakumar et al. 2004; Norenberg et al. 2004; Hertz and Kala 2007). Finally, astrocytes, but not neurones, exhibit considerable swelling under hyperammonaemic conditions, which is regarded as the main cause of fatal brain oedema following acute liver failure (Martinez 1968; Norenberg 1977; Norenberg et al. 2007; Vaquero and Butterworth 2007). The different susceptibility of astrocytes and neurones to increased NH4+/NH3 concentrations may be related to different influx and transport pathways across the plasma membrane for NH4+. However, despite the involvement of NH4+ influx in the recycling of neurotransmitters and the pathology of HE, the pathway(s) of NH4+ influx into neurones and astrocytes in situ is unknown and only speculated from data involving cultured cells (Liu et al. 2003; Titz et al. 2006; Kelly et al. 2009). In cultured hippocampal astrocytes, we previously showed that large increases in [Na+]i, mediated by activation of Na+, K+, Cl) co-transporter (NKCC1) accompany NH4+/NH3 evoked intracellular acidifications (Kelly et al. 2009). To directly compare these data to the situation in the intact tissue, in which astrocytes and neurones are densely packed and may interact through a narrow and diffusion-restricted extracellular space, we examined the pathways of cellular NH4+ influx in acutely isolated slices of the hippocampus, a wellestablished model system for the study of the acute effects of NH3/NH4+ on brain cells (see Felipo and Butterworth 2002). As the passage of NH4+/NH3 across cell membranes results in stereotypical changes in pHi, we monitored NH4+ influx by performing quantitative, ratiometric wide-field imaging of intracellular pH. In addition, quantitative sodium imaging was performed in both cell types to analyse Na+dependent transport processes related to NH4+/NH3.

Materials and methods Experimental tissue, solutions and test compounds The study and all experiments were carried out in accordance with the guidelines of the Heinrich-Heine-University Du¨sseldorf as well as the European Communities Council Directive (86/609/EEC) and approved by the institutional animal care and use committee.

Preparation of acute hippocampal slices from male Balb/c mice, bred and raised in the animal care facility of the Heinrich-HeineUniversity, was performed using standard techniques (see Meier et al. 2006; Kafitz et al. 2008). In brief, mice (postnatal days 19–22) were anaesthetised with carbon dioxide (CO2), decapitated and brains were rapidly placed in cold (4C) preparation ringer. Parasagittal slices (250 lm thick) were cut using a microtome (HM650 V; Microm International GmbH, Walldorf, Germany), incubated for 30 min at 34C in a holding chamber and loaded with sulforhodamine 101 (SR101) as previously described (Kafitz et al. 2008). Slices were then allowed to recover for at least 1 h at room temperature (19–23C) before being transferred as needed to a recording chamber, where they were superfused (2.5 mL/min) with solutions at 25C. Experiments were performed in standard HCO3)/CO2-buffered solution containing (in mM): NaCl 130, KCl 2.5, NaHCO3 26, NaH2PO4 1.25, MgSO4 1, CaCl2 2, D-glucose 20; pH 7.4 after equilibration with 5% CO2/95% O2. In nominally Cl)-free solutions with K+ channel blockers (low[Cl)]o/K+ solutions), gluconate salts replaced the respective chloride salts and TTX (0.5 lM), BaCl2 (1 mM) and TEA-Cl (2 mM) were added, resulting in a total [Cl)]o concentration of 4 mM. In nominally K+-free solutions, KCl was replaced with 2.5 mM NaCl or NH4Cl. Solutions containing 5 mM NH4Cl were prepared by equimolar substitution for NaCl. Solutions had a final osmolarity 300 ± 5 mOsm/L. Fluorescent ion imaging The movement of NH4+/NH3 across cell membranes results in stereotypical changes in pHi (see Boron 1983; Thomas 1984). Albeit an indirect measure, determination of NH4+/NH3-induced pHi changes has been shown to serve as an adequate tool for the study of mechanisms of its membrane passage (Nagaraja and Brookes 1998; Titz et al. 2006). Measurements of pHi and [Na+]i were performed on SR101 identified astrocytes and on CA1 pyramidal neurones in hippocampal slices. Standard dual excitation ratiometric measurements of 2¢,7¢-bis-(2-carboxyethyl)-5-(and-6)-carboxyfluorescein (for pH measurement) and sodium binding benzofuran isophthalate (for Na+ measurement) emission intensities were performed and converted to pHi and [Na+]i, respectively (see Data S1). Neurones and astrocytes were loaded with the acetoxymethyl ester forms of the fluorophores, pressure ejected via glass micropipettes placed in the stratum radiatum (Stosiek et al. 2003; Meier et al. 2006; Langer and Rose 2009). Following fluorophore application, hippocampal slices were incubated for 45 min at 19– 23C to allow the de-esterification of dyes. Data analysis The changes in steady-state pHi and [Na+]i evoked by a test manoeuvre were quantified as the respective difference between the mean steady-state pHi and [Na+]i values observed under the test condition with the mean of the steady-state pHi and [Na+]i values observed just before the test condition. Data are presented as mean ± SD, with the accompanying n-value referring to the number of cells from which data were obtained, and all experimental manoeuvres were performed on slices obtained from at least three different animals. Microspectrofluorimetric data were analysed in Igor Pro v.6 (Wavemetrics Inc., Lake Oswego, OR, USA) and Excel v.9 (Microsoft Corp., Redmond, WA,

 2010 The Authors Journal of Neurochemistry  2010 International Society for Neurochemistry, J. Neurochem. (2010) 115, 1123–1136

NH4+/NH3 influx into neurones and astrocytes | 1125

USA), and figures prepared using Adobe Illustrator CS2 (Adobe Systems, San Jose, CA, USA). Statistical comparisons were performed using Student’s two-tailed t-test, paired or unpaired as appropriate, with a 95% confidence limit.

Results NH4+/NH3 changes intracellular pH Initial pHi under control conditions was 7.22 ± 0.03 in SR101-positive astrocytes (n = 23) and 7.21 ± 0.04 in CA1 neurones (n = 43). Bath application of 5 mM NH4+/NH3 evoked transient alkalinisations, which in astrocytes reached peak amplitudes of 0.06 ± 0.04 pH units at 50 ± 15 s after the onset of the NH4+/NH3 application (Fig. 1a), and in neurones reached peak amplitudes of 0.05 ± 0.03 pH units

(a) pHi changes NH4+/NH3

6.9

6.9

7.0

7.0

7.2

7.2

5 min 7.3

7.3

23 23 + Control NH4 /NH3

NH /NH3

6.8

6.8

6.9

6.9

7.0

7.0

–0.27 ± 0.09

**

7.1

7.1

7.2

7.2

5 min

7.3

7.3

(b) [Na+]i changes Astrocytes

NH4+/NH3

40

40

[Na+]i (mM)

[Na+]i (mM)

50

30 20 10

**

30 20

13 0

NH4+/NH3

13

Control NH4+/NH3 1 ± 5 mM

[Na+]i (mM)

[Na+]i (mM)

43

10

5 min

0

20 10 0

43

Control NH4+/NH3

27 ± 8 mM

50

Neurones

NH4+/NH3-evoked [Na+]i increases accompany acidifications in astrocytes We previously showed that large increases in [Na+]i accompany NH4+/NH3 evoked intracellular acidifications in cultured astrocytes (Kelly et al. 2009). In the present study and in agreement with previous studies using cultured cells and slice preparations, steady-state [Na+]i was 13 ± 3.4 mM in astrocytes (n = 13) and 15 ± 2.7 mM in neurones (n = 14) under control conditions (see Rose and Ransom 1996a, 1997; Pisani et al. 1998; Calabresi et al. 1999; Chatton et al. 2000;

+ 4

pHi

pHi

**

7.1

7.1

Neurones

–0.18 ± 0.05

pHi

pHi

Astrocytes

after 50 ± 15 s (Fig. 1a). In the continuous presence of NH4+/NH3, alkalinisations were superseded by acidifications of 0.18 ± 0.05 and 0.27 ± 0.09 pH units in astrocytes and neurones, respectively (Fig. 1a). Whereas transient alkalinisations reflect direct passage of NH3 across the cell membrane, the large decreases in pHi likely reflect the influx of NH4+ into the cells, promoting the intracellular formation of NH3 and the release of H+ (see Fig. 6; Roos and Boron 1981; Boron 1983; Thomas 1984). Recovery from this acidification after withdrawal of NH4+/NH3 was reported to be largely mediated by Na+-dependent export of acid equivalents, specifically by Na+/H+-exchange in addition to activation of inwardly directed Na+ and HCO3)-dependent transporters (Deitmer and Rose 1996; Chesler 2003; Kelly et al. 2009). Upon removal of NH4+/NH3, rebound acidifications were observed, after which pHi again returned to near control values (Fig. 1a).

5 min

20 10 0

14

14

Control NH4+/NH3

Fig. 1 NH4+/NH3-evoked pHi and [Na+]i changes in astrocytes and neurones of hippocampal slices. (a) pHi changes in astrocytes (upper traces) and neurones (lower traces) in the same hippocampal slice induced by bath application of 5 mM NH4+/NH3 (indicated by the bars). Grey traces are experimental records from individual cells and the black trace represents the mean response of the displayed traces. Note that in this and subsequent figures, alkalinisations and acidifications are depicted as downward and upward deflections, respectively. Gaps in the traces represent pauses in data acquisition for approximately 10 min. Right panel: plots summarising the mean steady-state pHi + SD values in astrocytes (upper plot) and neurones (lower plot) observed prior to, and during prolonged NH4+/NH3 applications. n-Values are contained within the bars. **p < 0.01 for the difference between the indicated pHi value obtained prior to and in the presence of NH4+/NH3. (b) [Na+]i changes in astrocytes (upper traces) and neurones (lower traces) in the same hippocampal slice evoked by bath application of 5 mM NH4+/NH3 (indicated by the bars). Grey traces are experimental records from individual cells and the black trace represents the mean response of the displayed traces. Gaps in the traces represent pauses in data acquisition for approximately 10 min. Right panel: plots summarising the mean steady-state [Na+]i + SD values in astrocytes (upper plot) and neurones (lower plot) observed prior to and in the presence of NH4+/NH3. n-Values are contained within the bar. **p < 0.01 for the difference between the indicated [Na+]i value obtained prior to and in the presence of NH4+/ NH3.

 2010 The Authors Journal of Neurochemistry  2010 International Society for Neurochemistry, J. Neurochem. (2010) 115, 1123–1136

1126 | T. Kelly and C. R. Rose

(a)

Involvement of NKCC in pHi changes

Astrocytes/bumetanide 7.1

–0.10 ± 0.04

NH4+/NH3

7.1

7.2 7.3

7.3

7.4

7.4

5 min

6.9

NH4+/NH3

**

7.0 pHi

pHi

25 + NH4+

6.9

7.1 7.2

7.1 7.2

5 min

7.3

(b)

25 + Bum

–0.24 ± 0.09

Neurones/bumetanide

7.0

7.3

37 37

37 37

+ Bum

+ NH4+

Involvement of NKCC in [Na+]i changes

Astrocytes/bumetanide 30

NH4+/NH3 [Na+]i (mM)

[Na+]i (mM)

30 20 10

6 ± 5 mM

20 10

34

5 min

0

0

34

+ Bum + NH4+

Neurones/bumetanide NH4+/NH3

30 [Na+]i (mM)

30 [Na+]i (mM)

Fig. 2 Involvement of NKCC in NH4+/NH3-evoked acidifications and [Na+]i changes. All traces were obtained in the presence of the NKCC inhibitor bumetanide (100 lM). (a) pHi changes in astrocytes (upper traces) and neurones (lower traces) in the same hippocampal slice evoked by bath application of 5 mM NH4+/NH3 (indicated by the bars) in the presence of bumetanide. Grey traces are experimental records from individual cells and the black trace represents the mean response of the displayed traces. Gaps in the traces represent pauses in data acquisition for approximately 10 min. Right panel: plots summarising the mean steady-state pHi + SD values in astrocytes (upper plot) and neurones (lower plot) in the presence of bumetanide observed prior to, and during prolonged NH4+/NH3 applications. n-Values are contained within the bars. (b) [Na+]i changes in astrocytes (upper traces) and neurones (lower traces) in the same hippocampal slice evoked by bath application of 5 mM NH4+/NH3 (indicated by the bars). Grey traces are experimental records from individual cells and the black trace represents the mean response of the displayed traces. Gaps in the traces represent pauses in data acquisition for approximately 10 min. Right panel: plots summarising the mean steady-state [Na+]i + SD values in astrocytes (upper plot) and neurones (lower plot) observed prior to and in the presence of NH4+/NH3. n-Values are contained within the bars. **p < 0.01 for the difference between the indicated [Na+]i value obtained prior to and in the presence of NH4+/NH3.

**

7.2 pHi

Involvement of NKCC in NH4+/NH3-evoked changes in pHi and [Na+]i in astrocytes The NH4+/NH3-evoked steady-state acidifications and [Na+]i increases in astrocytes of hippocampal slices are similar to those observed in cultured astrocytes, which were mediated by NH4+ influx via activation of the Na+, K+, Cl) co-transporter (NKCC1; Kelly et al. 2009). Therefore, in the present study, we examined the contribution of NKCC activity to the NH4+/NH3 evoked steady-state acidifications and [Na+]i increases by applying the specific NKCC inhibitor bumetanide. Inhibition of NKCC differentially affected NH4+/NH3evoked ion changes in astrocytes and neurones of hippocampal slices. In the presence of bumetanide, NH4+/NH3 application still evoked acidifications in astrocytes

()0.10 ± 0.04 pH units; n = 25; Fig. 2a). However, bumetanide significantly (p < 0.01) reduced the amplitude of NH4+/NH3-evoked acidifications in astrocytes compared with acidifications observed in the absence of bumetanide (see Fig. 6a). In contrast, NH4+/NH3-evoked acidifications in neurones were not significantly altered in the presence of bumetanide (n = 37; Fig. 2a; see Fig. 6a). Additionally, we examined the involvement of NKCC in NH4+/NH3-evoked [Na+]i changes. In astrocytes, although transient NH4+/NH3-evoked [Na+]i decreases were still observed, bumetanide prevented NH4+/NH3-evoked increases in steady-state [Na+]i (6 ± 5 mM; n = 34; Fig. 2b; cf. Fig. 1b). Similar to results obtained in the absence of bumetanide, NH4+/NH3 failed to elicit steady-state [Na+]i changes in neurones in the presence of bumetanide (Fig. 2b; n = 28). Therefore, in contrast to astrocytes, NKCC inhibition failed to significantly alter the effects of NH4+/NH3 on steady-state pHi and [Na+]i in neurones of hippocampal slices (see Fig. 6).

pHi

Sheldon et al. 2004; Langer and Rose 2009). Application of NH4+/NH3 significantly increased [Na+]i in astrocytes. Similar to findings in cultured astrocytes, initial transient decreases in [Na+]i (Fig. 1b) of 3 ± 1 mM after 199 ± 2 s were followed by large increases in [Na+]i to new steadystate values (n = 13) within 20–30 min of continuous NH4+/ NH3 application. The mean amplitude of [Na+]i increases was 27 ± 8.0 mM, and the elevated [Na+]i was maintained in the presence of NH4+/NH3 (Fig. 1b). In contrast to astrocytes, NH4+/NH3 failed to significantly affect steady-state [Na+]i in neurones. In neurones, application of NH4+/NH3 transiently reduced [Na+]i by 2 ± 1 mM after 209 ± 25 s (Fig. 1b; p < 0.05), which was followed by a return of steady-state [Na+]i to near control values in the continuous presence of NH4+/NH3 (n = 14).

20 10 0

5 min

0.1 ± 0.1 mM

20 10 0

28

28

+ Bum

+ NH4+

 2010 The Authors Journal of Neurochemistry  2010 International Society for Neurochemistry, J. Neurochem. (2010) 115, 1123–1136

NH4+/NH3 influx into neurones and astrocytes | 1127

The above data suggest that the mechanisms of NH4+ influx differ between astrocytes and neurones. Similar to results in cultured astrocytes, bumetanide sensitive NKCC is at least partially involved in NH4+ influx in astrocytes of hippocampal slices. In contrast to hippocampal astrocytes, the data suggest NKCC is not involved in the influx of NH4+ into neurones in hippocampal slices. Cation chloride co-transport is not involved in NH4+/NH3-evoked pHi changes in neurones The different mechanisms of NH4+ influx between astrocytes and neurones are consistent with the differences in the expression of cation chloride co-transporters. Whereas rodent astrocytes express high levels of NKCC1 at around P20, hippocampal principle neurones mainly express K+, Cl) co-transporter (KCC2) and only minimal levels of NKCC at this stage of development (Lu et al. 1999; Yan et al. 2001). In light of these findings, we examined the involvement of KCC in the influx of NH4+.

Furosemide, an analogue of bumetanide, which inhibits both NKCC and KCC activity, was applied to hippocampal slices and NH4+/NH3-evoked pHi changes assessed. Similar results were observed in the presence of 100 lM and 1 mM furosemide, and data were pooled. Similar to the effects of bumetanide, furosemide reduced NH4+/NH3-evoked acidifications in astrocytes by about 55% (n = 16; Fig. 3). The reduction of NH4+/NH3-evoked acidifications by furosemide is consistent with furosemide inhibiting NKCC activity (see Payne 1997; Russell 2000), and complements the results obtained with bumetanide. In contrast, furosemide failed to reduce NH4+/NH3-evoked acidifications in neurones (n = 35; Fig. 3), suggesting that neither NKCC nor KCC activity were involved. To further investigate the possible involvement of cation chloride co-transport in the influx of NH4+, we perfused hippocampal slices with low [Cl)]o solutions to inhibit chloride-dependent transport. In addition, K+ channel inhibitors TEA (2 mM) and Ba2+ (1 mM BaCl2) were added to

Involvement of Cl–-dependent transport and K+-channels on pHi changes Astrocytes/furosemide – Furosemide

low[Cl ]o + Ba2+ + TEA 6.9

6.9

+

NH4 /NH3

7.0

7.2

7.1

pHi

pHi

7.1

7.3

–0.12 ± 0.04

–0.08 ± 0.03

7.0 7.1

**

7.2

pHi

7.0

7.3

7.3 7.4

16

7.4

5 min

7.5

Neurones/furosemide 6.8

16

6.9

7.4

+ Furos + NH4+

Furosemide

NH4+/NH3

*

7.2

–0.35 ± 0.10

6.8

**

7.0

13

low[Cl–]o + Ba2+ + TEA

–0.22 ± 0.06

6.9

13

low[Cl–]o/ + NH4+ K+blockers

**

6.9 7.0

7.0 7.1

pHi

pHi

pHi

7.1 7.1

7.2 7.2 7.3

7.2 7.3

5 min

Fig. 3 Involvement of Cl)-dependent transporters and K+ channels on NH4+/NH3-evoked pHi changes. All traces were obtained in the presence of the KCC inhibitor furosemide (100 lM). Upper left panel: pHi changes evoked by NH4+/NH3 in astrocytes. Grey traces are experimental records from individual cells and the black trace represents the mean response of the displayed traces. Gaps in the pHi traces represent pauses in data acquisition for approximately 10 min. Right panels: plots summarising the mean steady-state pHi + SD values in astrocytes under the indicated conditions observed prior to, and during NH4+/NH3 applications. *p < 0.05 and **p < 0.01 for the difference

7.4

7.3 35 35

35 35

+ Furos + NH4+

7.4

25

25

low[Cl–]o/ + NH4+ K+blockers

between the indicated pHi value obtained prior to and in the presence of NH4+/NH3. n-Values are contained within the bars. Lower left panel: pHi changes evoked by NH4+/NH3 in neurones. Grey traces are experimental records from individual cells and the black trace represents the mean response of the displayed traces. Right panels: plots summarising the mean steady-state pHi + SD values in neurones under the indicated conditions observed prior to, and during NH4+/NH3 applications. n-Values are contained within the bars. **p < 0.01 for the difference between the indicated pHi value obtained prior to and in the presence of NH4+/NH3.

 2010 The Authors Journal of Neurochemistry  2010 International Society for Neurochemistry, J. Neurochem. (2010) 115, 1123–1136

1128 | T. Kelly and C. R. Rose

Ouabain inhibits NH4+/NH3-evoked pHi changes in astrocytes and neurones The application of 5 lM ouabain failed to significantly change steady-state pHi in astrocytes (n = 18; see Fig. 4a). In neurones, in contrast, pHi significantly decreased from 7.22 ± 0.05 in the absence to 7.13 ± 0.08 in the presence of 5 lM ouabain (p < 0.05; n = 48). In the presence of ouabain, membrane integrity was assessed using the isosbestic point of 2¢,7¢-bis-(2-carboxyethyl)-5-(and-6)-carboxy-

Effects of ouabain on pHi changes

Astrocytes 7.1

–0.07 ± 0.03

Ouabain

7.1

+

NH4 /NH3

7.2 pHi

pHi

7.3

7.3

5 min

7.4

Neurones

7.4

Ouabain

7.0

7.0

pHi

pHi

5 min

7.3

7.2 7.3

Effects of ouabain on [Na+]i changes

6 ± 5 mM

*

40

NH4+/NH3 [Na+]i (mM)

[Na+]i (mM)

48 48 + Ouabain + NH4+

Ouabain

40 30 20

5 min

10

30 20 10 0

Neurones

Ouabain

16

16

+ Ouabain + NH4+ 5 ± 3 mM

NH4+/NH3

*

30

30 20 10

18

7.1

7.2

Astrocytes

18

+ Ouabain + NH4+ –0.01 ± 0.05

NH4+/NH3

7.1

(b)

*

7.2

[Na+]i (mM)

Involvement of Na+, K+-ATPase in NH4+/NH3-evoked pHi changes Previous studies suggested an alternate NH4+ influx pathway via Na+, K+-ATPase in invertebrate neurones and rat kidney collecting duct cells (Moser 1987; Wall 1997). In crayfish neurones, NH4+ substitutes for K+ and is actively transported into the cell, resulting in decreases in pHi (Moser 1987). An influx of NH4+ via Na+, K+-ATPase may be responsible for the Cl)-independent portion of NH4+/NH3evoked acidifications observed in astrocytes and neurones in hippocampal slices (see above). Therefore, we investigated the contribution of Na+, K+-ATPase to NH4+/NH3-evoked acidifications in our preparation. The specific Na+, K+ATPase inhibitor ouabain was employed at 5 lM to block the a2, astrocyte-specific, and a3, neurone-specific, subunits; whereas the ubiquitous a1 subunit is relatively insensitive to ouabain at this concentration (IC50 50 lM; O’Brien et al. 1994).

(a)

[Na+]i (mM)

low [Cl)]o solutions to inhibit NH4+ influx through K+ channels. Inhibition of Cl)-dependent transport and K+ channels reduced NH4+/NH3-evoked acidifications in astrocytes (n = 13; Fig 3). The amplitudes of NH4+/NH3-evoked acidifications observed with low [Cl)]o/K+ channel blockers were not significantly different from the amplitudes obtained in the presence of furosemide or bumetanide (see Fig. 6a). In contrast to astrocytes, low [Cl)]o/K+ channel blockers failed to reduce NH4+/NH3-evoked acidifications in neurones (n = 25; Fig. 3). Instead, the amplitude of acidifications observed in neurones during NH4+/NH3 was increased under these conditions (see Fig 6b). The lack of an additional effect on the NH4+/NH3-evoked steady-state acidifications in astrocytes by furosemide or low [Cl)]o solution/K+ channel blockers compared with bumetanide, suggests that each of these manoeuvres inhibited NH4+ influx primarily via inhibition of NKCC. However, inhibition of NKCC with bumetanide or furosemide reduced the amplitude of NH4+/NH3-evoked acidifications in astrocytes by only 45–55%, suggesting that additional mechanisms contribute to NH4+ influx in astrocytes. Moreover, and contrary to previous studies in cultured neurones (Liu et al. 2003; Titz et al. 2006), the data suggests that Cl)-dependent transport and K+ channels are not the primary mechanisms for NH4+ influx in neurones of hippocampal slices.

20 10

5 min

26 0

26

+ Ouabain + NH + 4

Fig. 4 Effects of ouabain on NH4+/NH3-evoked pHi and Na+i changes. (a) pHi traces from astrocytes (upper traces) and neurones (lower traces) in a hippocampal slice. Application of ouabain (5 lM) and NH4+/NH3 (5 mM) is indicated by the bars. Grey traces are experimental records from individual cells and the black trace represents the mean response of the displayed traces. Gaps in the traces represent pauses in data acquisition for approximately 10 min. Right panels: plots summarising the mean steady-state pHi + SD values in the presence of ouabain in astrocytes (upper plot) and neurones (lower plot) prior to, and during NH4+/NH3 applications. *p < 0.05 for the difference between the indicated pHi value obtained prior to and in the presence of NH4+/NH3. n-Values are contained within the bars. (b) [Na+]i traces from astrocytes (upper traces) and neurones (lower traces) in a hippocampal slice. Application of ouabain (5 lM) and NH4+/ NH3 (5 mM) is indicated by the bars. Grey traces are experimental records from individual cells and the black trace represents the mean response of the displayed traces. Gaps in the traces represent pauses in data acquisition for approximately 10 min. Right panels: plots summarising the mean steady-state [Na+]i + SD values in the presence of ouabain in astrocytes (upper plot) and neurones (lower plot) prior to, and during NH4+/NH3 applications. n-Values are contained within the bars. *p < 0.05 for the difference between the indicated [Na+]i value obtained prior to and in the presence of NH4+/NH3. n-Values are contained within the bars.

 2010 The Authors Journal of Neurochemistry  2010 International Society for Neurochemistry, J. Neurochem. (2010) 115, 1123–1136

NH4+/NH3 influx into neurones and astrocytes | 1129

fluorescein (see Bevensee et al. 1995), and cells in which the 452 nm fluorescence intensity decreased by more than 5% per minute were discarded. In both astrocytes and neurones, ouabain significantly reduced NH4+/NH3-evoked steady-state acidifications (see Fig. 4a). Following 20 min in the presence of ouabain, the subsequent application of NH4+/NH3 elicited transient alkalinisations that were superseded by steady-state acidifications in astrocytes (Fig. 4a). Although the NH4+/NH3-evoked transient alkalinisations were similar to those observed in the absence of ouabain, the NH4+/NH3-evoked steady-state acidifications were significantly reduced in the presence compared with the absence of ouabain. Thus, NH4+/NH3 changed pHi in astrocytes by )0.07 ± 0.03 pH units in the presence of ouabain, a 61% reduction compared with control conditions (see above; Fig. 6a). In neurones, ouabain prevented NH4+/NH3-evoked acidifications. Thus, in the presence of ouabain, pHi in neurones prior to and during NH4+/NH3 was not significantly different. The NH4+/NH3-evoked pHi changes in ouabain ()0.01 ± 0.05, n = 48; Fig. 4a) were 96% reduced compared with NH4+/NH3-evoked acidifications observed in its absence (see Fig. 6b). The strong decline of NH4+/NH3-evoked acidifications observed with ouabain might be secondary to an increase in intracellular Na+, reducing the driving force for Na+-dependent NH4+ influx. We thus determined ouabain-induced [Na+]i changes in astrocytes and neurones. Additionally, we examined the effects of Na+, K+-ATPase inhibition on NH4+/ NH3-evoked Na+ changes. Application of 5 lM ouabain increased [Na+]i in both astrocytes and neurones as expected following Na+, K+-ATPase inhibition. The ouabain-induced [Na+]i increases, initiated immediately following ouabain application, began to reach a plateau after 20 min (Fig. 4b). In astrocytes, [Na+]i increased from 13 ± 2 mM under control conditions to 29 ± 5 mM in the presence of ouabain (n = 16; p < 0.01; Fig. 4b). In neurones, [Na+]i increased from 11 ± 4 mM under control conditions to 20 ± 6 mM in the presence of ouabain (n = 26; p < 0.01; Fig. 4b), and this increase by 9 ± 5 mM was significantly smaller (p < 0.01) than that in astrocytes (16 ± 5 mM). In addition, in the presence of ouabain the NH4+/NH3-evoked [Na+]i increases (6 ± 5 mM; n = 16) in astrocytes were significantly (p < 0.01) reduced compared with those observed in the absence of ouabain (Fig. 4b; cf. Fig. 1b). In neurones, although NH4+/NH3 increased [Na+]i in the presence of ouabain, the amplitude of the NH4+/NH3-evoked changes in steady-state [Na+]i were not significantly different in the presence of ouabain compared with the absence (p = 0.15; n = 26 and 14, respectively). Taken together these data show that ouabain significantly reduces NH4+/NH3-evoked acidifications in astrocytes, whereas it nearly blocks NH4+/NH3-evoked acidifications in neurones, indicating that Na+, K+-ATPase activity is an

important pathway for NH4+ influx. However, because at the same time ouabain increased [Na+]i in both cell types, these effects might also have been caused by the decreased driving force for Na+ instead of the inhibition of Na+, K+-ATPase activity itself. We, therefore, performed further experiments to study if Na+, K+-ATPase can transport NH4+. NH4+ substitutes for K+ at Na+, K+-ATPase As mentioned above, the ionic radius of NH4+ is similar to K+, and NH4+ has been shown to effectively substitute for K+ at many transporters including the Na+, K+-ATPase (Knepper et al. 1989). Biochemical in vitro studies in invertebrates found that removal of K+ inhibits Na+, K+-ATPase activity, and that NH4+ effectively substitutes for the absence of K+ (see Skou 1960; Towle 1984). Furthermore, these studies suggested that the half maximal effective concentration of NH4+ at the Na+, K+-ATPase was about three to five times that of K+. To assess whether NH4+ substitutes for K+ at Na+, K+-ATPase in hippocampal neurones, we examined the effects of NH4+ on the [Na+]i increases observed following inhibition of Na+, K+-ATPase with nominally [K+]o-free solution. We chose to examine neurones as our data suggest that Na+, K+-ATPase activity was the primary mechanism for NH4+ influx into neurones. The large increases in [Na+]i evoked by NH4+ via NKCC activity in astrocytes preclude a straight forward interpretation of how NH4+ effects [Na+]i increases induced by nominally [K+]o-free solution. Removal of extracellular K+ increases [Na+]i in neurones, consistent with the inhibition of Na+, K+-ATPase and accumulation of intracellular Na+ (e.g. see Rose and Ransom 1997). Exposure of CA1 neurones to nominally [K+]o-free solution for 5 min continuously increased [Na+]i from 15.8 ± 3.8 to 23.3 ± 4.7 mM (p < 0.01; n = 39; Fig. 5a). The rate of [Na+]i increases changed during the continuous exposure to nominally [K+]o-free solution; decreasing from 2.2 ± 0.8 mM/min in the first minute to 0.8 ± 0.3 mM/min during the final minute of nominally [K+]o-free solution (p < 0.01; n = 39; see Fig. 5). The kinetics of the induced [Na+]i increases in slices differ to the linear increase observed in cultured hippocampal neurones (Rose and Ransom 1997), and may reflect the slower and incomplete removal of extracellular K+ in slices. Replacement of extracellular K+ with NH4+ attenuated the amplitude of [Na+]i increases induced by nominally [K+]ofree solution by 42 ± 10% ([Na+]i increased by 7.5 ± 1.5 mM in 0[K+]o compared with 4.3 ± 1.4 mM in 0[K+]o/ 2.5 mM [NH4+]o; p < 0.01; n = 39, Fig. 5). Furthermore, NH4+ substituted solutions altered the kinetics of the [Na+]i transients induced by nominally [K+]o-free solution. Although the initial rate of increase was similar between 0[K+]o and NH4+ substituted solutions, NH4+ substitution reduced the rate of Na+ increases during the final minute by 88 ± 37%. Therefore, and in contrast to the continuous

 2010 The Authors Journal of Neurochemistry  2010 International Society for Neurochemistry, J. Neurochem. (2010) 115, 1123–1136

1130 | T. Kelly and C. R. Rose

(a)

slices (Fig. 6). In astrocytes of hippocampal slices, NH4+ influx was partly mediated by Na+, K+-ATPase activity and partly by NKCC activity, which also caused large increases in [Na+]i. In CA1 neurones of hippocampal slices, NH4+ primarily entered via Na+, K+-ATPase activity, and our data are consistent with NH4+ substituting for K+.

NH4+ replaces K+ at the Na+,K+-ATPase +

0[K ]o

0[K+]o

0[K+]o

+ NH4+/NH3

[Na+]i (mM)

30 25

Discussion

20 15 5 min 0[K+]o

(c)

5 mM

5 min

+ NH4+

0[K+]o-induced [Na+]i increase (mM)

(b)

10

–42 ± 10%

8

**

6

The results of the present study provide mechanisms for the influx of NH4+ into astrocytes and neurones in hippocampal slices following rapid elevation of NH4+, relevant during the onset of acute hyperammonaemic conditions. Our data indicate that both NKCC1 and Na+, K+-ATPase activity contribute to the influx of NH4+ into astrocytes of hippocampal slices. In neurones of hippocampal slices, ouabain effectively blocked influx of NH4+, indicating a major role for Na+, K+-ATPase activity.

4 2 0

39

39

– NH4+ + NH4+

Fig. 5 NH4+ reduces [Na+]i increases induced by nominally [K+]o-free solution in neurones. (a) [Na+]i traces from neurones. Grey traces are experimental records from individual cells and the black trace represents the mean response of the displayed traces. Removal of extracellular K+ (0[K+]o) for 5 min reversibly increased [Na+]i. Replacement of K+ with 2.5 mM NH4+ reversibly reduced the amplitude of 0[K+]oinduced [Na+]i increases. (b) Grey traces: mean [Na+]i responses to 0[K+]o in the presence and absence of 2.5 mM NH4+/NH3 shown in (a) superimposed to aid comparison. Black curves: simulation of experimental data using a simple mathematical model according to Takeuchi et al. (2006; see Data S1). Reducing extracellular [K+] to 0.4 mM in the model simulated the experimentally measured [Na+]i increase in 0[K+]o. Replacing [K+]o with NH4+, which was given a Kd (1 mM) four times greater than that of K+ (0.25 mM) at the Na+, K+-ATPase, simulated the experimentally measured [Na+]i increase in NH4+ substituted 0[K+]o solutions. (c) Plot summarising the mean 0[K+]o-induced [Na+]i increase in the presence and absence of substituted NH4+. **p < 0.01 for the difference between the indicated 0[K+]o-induced [Na+]i increase obtained in the presence of substituted NH4+ and the corresponding value obtained in the absence of NH4+.

increases in [Na+]i during nominally [K+]o-free solution, [Na+]i reached a new steady-state value in 0[K+]o/2.5 mM [NH4+]o. These data are consistent with NH4+ replacing K+, albeit less effectively, at the Na+, K+-ATPase. In fact, Na+ increases observed in 0[K+]o/2.5 mM [NH4+]o were mathematically simulated by a NH4+ Kd 4 times higher than that of K+ at the K+ binding site (see Fig. 5b; see Data S1). In conclusion, our data suggest that the mechanisms of NH4+ influx and the consequent intracellular ion changes differ between astrocytes and CA1 neurones of hippocampal

NH4+ influx into astrocytes The influx of NH4+ into cells results in intracellular acidifications in a variety of cell types (see Boron 1983; Thomas 1984; Marcaggi and Coles 2001) and, moreover, pHi measurements have been successfully employed to characterise the mechanisms employed by astrocytes and neurones to transport NH4+ (Nagaraja and Brookes 1998; Titz et al. 2006). Previous data from cultured cortical and hippocampal astrocytes showed NH4+/NH3-evoked steadystate acidifications because of the transport of NH4+ into the cell (Nagaraja and Brookes 1998; Titz et al. 2006; Kelly et al. 2009). However, reports conflict with regard to the significance of the NH4+ influx pathways, some reports suggest that NKCC1 activity is the predominant influx mechanism (Titz et al. 2006; Kelly et al. 2009); whereas other reports suggest NKCC1 activity plays a minor role relative to inward rectifying K+ (Kir) channels (Nagaraja and Brookes 1998). These differences may result from culture conditions and the maturation level of cultured astrocytes (see Kelly et al. 2009). Consistent with the age of animals employed in the present study (P19–22), we found that NKCC contributed to the influx of NH4+ into astrocytes in hippocampal slices. However, inhibition of NKCC activity with bumetanide only partially (50%) reduced the NH4+/NH3-evoked acidification influx in hippocampal astrocytes in slices. As an additional mechanism to inhibit NKCC, we employed low [Cl)]o solutions, which contained TEA and Ba2+to block a variety of K+ channels including Kir channels. Similar to observations in the presence of bumetanide, NH4+/NH3evoked acidifications were reduced but not prevented by the low [Cl)]o/K+ solutions. Both the data obtained in the presence of bumetanide and in low [Cl)]o/K+ solutions are consistent with the involvement of NKCC and an additional Cl)-independent mechanism in the influx of NH4+ into

 2010 The Authors Journal of Neurochemistry  2010 International Society for Neurochemistry, J. Neurochem. (2010) 115, 1123–1136

NH4+/NH3 influx into neurones and astrocytes | 1131

Summary of NH4+/NH3-influx pathways in astrocytes and neurones (a) Astrocytes

NH4+/NH3-evoked acidification

–0.25 K+

–0.20

Cl–

**

**

–0.15

NH4+

*

–0.10

**

NH4+

NH4+ H+

H+

ADP+Pi

+

H NH3

NH3

3 Na+

2 NH4+

ATP

–0.05 23

25

16

13

Cl– Na+ NH4+

18

b

2 Cl– NH4+

K+

+O ua

[C blo l –] / cke o rs

Co ntr ol +B um +F ur low os

0.00

(b) Neurones NH4+

– NH4+ Cl

–0.3

NH4+

NH4+ H+

H+

–0.2

H+ NH3

NH3

ADP+Pi

2 NH4+

3 Na+ ATP

–0.1

**48

25

Cl–

2 Cl–

Na+

+

NH4+

NH4

Ou

uro

ab

35

s lo K + w[C – l blo ] / cke o rs +

+B

Co

ntr

ol

0.0

37

um

47

+F

NH4+/NH3-evoked acidification

–0.4

Fig. 6 Summary of NH4+/NH3 influx pathways in astrocytes and neurones. (a) Bar chart summarising the mean peak amplitudes of the NH4+/NH3-evoked steady-state acidifications under the indicated conditions in astrocytes. ‘Control’ indicates the situation under initial conditions, the other bars reflect the difference between the antagonist alone and the antagonist with NH4+/NH3. Right panel: schematic of possible NH4+ influx pathways in astrocytes. The data support NH4+ influx mediated via NKCC and Na+, K+-ATPase activity in astrocytes of hippocampal slices. (b) Bar chart summarising the mean peak amplitudes of the NH4+/NH3-evoked steady-state acidifications under

the indicated conditions in neurones. ‘Control’ indicates the situation under initial conditions, the other bars reflect the difference between the antagonist alone and the antagonist with NH4+/NH3. Right panel: schematic of possible NH4+ influx pathways in neurones. The data support NH4+ influx mediated primarily via Na+, K+-ATPase activity in neurones of hippocampal slices. n-Values are contained within the bars. *p < 0.05 and **p < 0.01 for the difference between the NH4+/ NH3-evoked steady state acidification obtained under the indicated condition compared with the corresponding response obtained under control conditions.

hippocampal astrocytes. Thus, NKCC is an important mechanism for NH4+ influx into astrocytes of hippocampal slices. An additional Cl)-independent mechanism also exists and in contrast to some reports in cultured cells (Nagaraja and Brookes 1998) appears not to involve K+ channels. Previous studies suggested NH4+ influx was mediated by Na+, K+-ATPase activity (Moser 1987). We, therefore, investigated the involvement of Na+, K+-ATPase in the influx of NH4+ in hippocampal astrocytes using the selective inhibitor ouabain. A ouabain concentration of 5 lM was utilised to block both a2 and a3 isoforms of the Na+ pump

(IC50 = 59 and 7 nM, respectively; O’Brien et al. 1994), which are expressed in hippocampal astrocytes and neurones, respectively (Blanco and Mercer 1998). Consistent with the inhibition of Na+ pump activity, steady-state [Na+]i increased in both neurones and astrocytes and reached a new stable level of 30 and 20 mM in astrocytes and neurones, respectively, which was maintained in the presence of 5 lM ouabain. In contrast, studies employing high concentrations (1 mM) of ouabain, which inhibit all three isoforms, observed larger (‡ 30 mM within 10 min) and continuous [Na+]i increases (see Rose

 2010 The Authors Journal of Neurochemistry  2010 International Society for Neurochemistry, J. Neurochem. (2010) 115, 1123–1136

1132 | T. Kelly and C. R. Rose

and Ransom 1996a, 1997; Monteith and Blaustein 1998; also see Dobretsov and Stimers 2005). Low concentrations of ouabain (5–10 lM) have also been reported to only slightly increase extracellular potassium concentration in hippocampal slices (by 0.5–1 mM; see McCarren and Alger 1987; N. Haak, P. Hochstrate, C. R. Rose, unpublished data). Our results, therefore, are consistent with the inhibition of ouabain-sensitive a isoforms and the continuous activity of the ubiquitously expressed a1 isoform, which is relatively insensitive to ouabain (IC50 = 50 lM). However, a partial inhibition of the a1 isoform cannot be excluded. In astrocytes, inhibition of Na+, K+-ATPase activity reduced NH4+/NH3-evoked acidification, consistent with the involvement of the Na+, K+-ATPase activity in NH4+ influx. The 61% reduction of the NH4+/NH3-evoked acidification is likely an overestimation of the contribution of Na+, K+, ATPase activity as the increased steady-state [Na+]i in the presence of ouabain also likely reduced NKCC activity. Inhibition of NKCC by bumetanide, however, only partially reduced the acidifications (Fig. 2), indicating that ouabain might act by an additional mechanism, specifically by direct inhibition of NH4+-transport by the Na+, K+ATPase. However, the ouabain-induced Na+ increases may modulate pHi regulating mechanisms. Because the main Na+dependent pHi regulating mechanisms are acid extruding, that is, Na+/H+ exchange and Na+-dependent HCO3) cotransport (Roos and Boron 1981; Deitmer and Rose 1996), they, however, would be inhibited during elevated [Na+]i resulting in a larger, but not smaller NH4+/NH3 acidification. Astrocytes depolarise in the presence of ouabain (Walz and Wuttke 1999), which might alter NH4+ influx (and the resulting acidification) through ion channels, but we did not find convincing evidence for a significant contribution of such pathways (cf. Fig. 3). It has to be emphasised that our data do not exclude the possibility that NH4+/NH3 also enters astrocytes through aquaporins (Holm et al. 2005) or Rh glycoproteins (Huang and Ye 2010). In conclusion, based on our pharmacological experiments and on our evidence that NH4+ is efficiently transported by the Na+, K+, ATPase, we conclude that both NKCC and Na+, K+, ATPase activity are the predominant mechanisms that mediate the NH4+ influx into hippocampal astrocytes (Fig. 6a). NH4+ influx into neurones In cultured mammalian neurones, acidifications mediated by NH4+ influx were reduced by inhibition of K+, Cl) cotransport (KCC) activity (Liu et al. 2003; Titz et al. 2006). However, in the present study, inhibition of KCC activity with furosemide failed to reduce NH4+/NH3-evoked acidifications in hippocampal neurones. The inability of furosemide to inhibit the acidifications in neurones was unlikely caused by ineffective penetration of furosemide into hippocampal

slices, as furosemide, known to also inhibit NKCC activity, reduced the NH4+/NH3-evoked acidifications in astrocytes. Furthermore, low [Cl)]o solution containing K+ channels blockers also failed to reduce NH4+/NH3-evoked acidifications in hippocampal neurones. The above data obtained from neurones in hippocampal slices suggests that Cl)dependent transport activity and K+ channels were not significantly involved in the NH4+ influx (see Fig 6b) and contrasts with previous studies employing cultured mammalian neurones (Liu et al. 2003; Titz et al. 2006). In addition, the dedicated NH4+/Cl) co-transporter found in bee glial cells (Marcaggi et al. 1999) likely does not participate in NH4+ influx in hippocampal neurones. Interestingly, the absence of a dedicated NH4+ transport mechanism suggests that at physiological concentrations of NH4+/NH3 the influx of NH3 via Rh proteins and/or aquaporins may contribute to intracellular NH4+ accumulation (Holm et al. 2005; Huang and Ye 2010). We next investigated the involvement of Na+, K+ATPase activity in the NH4+ influx into hippocampal neurones. As mentioned above, the concentration of ouabain utilised in the present study predominantly inhibited the a3 and not the a1 Na+, K+-ATPase isoforms known to be expressed in CA1 pyramidal neurones (McGrail et al. 1991; also see Dobretsov and Stimers 2005). In the presence of 5 lM ouabain, NH4+/NH3 failed to evoked steady-state acidifications in CA1 neurones. Because NH4+ influx was not accompanied by a Na+ influx in neurones (cf. Fig. 1), it seems unlikely that the decreased driving force for Na+ in the presence of ouabain was the predominant mechanism responsible for the diminution of the acidification. As for astrocytes (see above), elevation of [Na+]i should result in an inhibition of the main acid extrusion mechanisms, promoting an increased acidification in the presence of ouabain. Also, we found no evidence for an involvement of ion channels in the NH4+-induced acidification (cf. Fig. 3). Thus, the sole manoeuvre to affect NH4+-evoked acidifications was inhibition of Na+, K+-ATPase activity. Notably, inhibiting Na+, K+-ATPase activity not only reduced, but in fact abolished NH4+-evoked acidifications. Taken together, this suggests that NH4+ influx was predominately mediated by the Na+ pump in CA1 neurones of hippocampal slices (see Fig. 6b). This argumentation is strongly supported by our experiments in which we assessed whether NH4+ acts as a K+ congener at the K+ binding site of the Na+, K+-ATPase by examining the effects of replacing K+ with NH4+ on [Na+]i increases induced by nominally K+-free solution. In agreement with previous in vitro biochemical studies and electrophysiological studies of Na-pump activity in invertebrate neurones and mammalian collecting duct cells (Towle 1984; Moser 1987; Wall 1997), the attenuation of [Na+]i increases induced by nominally K+-free solution by NH4+ substitution suggests that NH4+ replaces K+ at the Na+, K+-

 2010 The Authors Journal of Neurochemistry  2010 International Society for Neurochemistry, J. Neurochem. (2010) 115, 1123–1136

NH4+/NH3 influx into neurones and astrocytes | 1133

ATPase. This conclusion is also supported by a mathematical model of Na+, K+-ATPase function (see Takeuchi et al. 2006; Data S1), in which we assigned NH4+ a Kd value four times greater than K+ at the K+-binding site, and which replicated our experimental observations. In conclusion, our data suggest that NH4+ acts as a K+ congener at the K+ binding site of the Na+, K+-ATPase, and that Na-pump activity mediates NH4+ influx into CA1 neurones of hippocampal slices.

Conclusions Elevations in the NH4+ concentrations affect numerous cellular processes many of which are intracellular, and together these actions may contribute to the neurological deficits associated with HE. Our study provides mechanisms for NH4+ influx into neurones and astrocytes in hippocampal slices derived from juvenile mice. The hippocampus is a well-established model system for the study of the acute effects of NH3/NH4+ on brain cells, and among others, these studies revealed that elevation of NH3/NH4+ severely disturbs glutamatergic transmission in the hippocampus and other regions of the CNS (cf. Table 1 and references in Felipo and Butterworth 2002). Because astrocytes may exhibit region-specific heterogeneity (Zhang and Barres 2010), it may be informative to carry out similar experiments in the neocortex, where astrocytes have been reported to gain a reactive phenotype (‘Alzheimer Type II’ astrocytes) during prolonged elevation of NH3/NH4+ (Norenberg and Lapham 1974; see also review by Felipo and Butterworth 2002). Likewise, it has to be kept in mind that differences between different animal model systems (e.g. rat vs. mice) may exist. Although mice offer the advantage of generating relevant knockout/transgenic animals, rats are better suited for portocaval anastomosis, a widely used model for experimental chronic HE. In the present study, intracellular acidifications due to NH4+ influx were observed in both hippocampal neurones and astrocytes. The NH4+ induced acidifications are similar in amplitude to those observed during anoxia (Sheldon et al. 2004) and those that lead to mitochondial dysfunction and swelling (Ding et al. 2000; Schneider et al. 2004 and see Casey et al. 2010), and may contribute to the detrimental effects of NH4+. Changes in pHi are known to affect Ca2+release from intracellular stores (see Thomas 2002) an important signalling mechanism in both astrocytes and neurones (see Berridge 1998; Fiacco and McCarthy 2006). The initial transient alkalinisation caused by NH3 influx results in [Ca2+]i increases and consequently Ca2+-dependent vesicular release of glutamate (see Rose et al. 2005; Go¨rg et al. 2010). However, the subsequent prolonged decreases in pHi likely reduce intracellular Ca2+release in response to neurotransmitters and may add to the deficits in both glutamatergic and GABAergic neurotransmission observed

in HE patients (Monfort et al. 2002; Ahboucha and Butterworth 2004). The susceptibility of astrocytes to NH4+/NH3 compared with neurones is likely in large part because of the location of astrocytic end-feet lining blood vessels and astrocytes being the site of NH4+/NH3 entry into the CNS. The blood–brain barrier is the central cellular element that defines the entry of ammonia into the brain (Lockwood 2004). Although ammonia levels linearly correlate with blood ammonia levels over a wide concentration range, ammonia distribution is also governed by the pH gradient when considering extra- and intracellular compartments (Hindfelt 1975). In this regard, it was also found that the pH gradient between blood and brain is a major determinant for the flux of ammonia over the blood–brain barrier, suggesting that mild metabolic acidosis might be helpful in treating hyperammonaemia (Lockwood et al. 1980). Considering the last cellular element of the blood–brain barrier, the astrocyte endfeet, it is interesting to note that, expression of NKCC1 is localised to astrocytic end-feet (Yan et al. 2001) and our data suggest that NKCC may contribute to the loading of NH4+ into the CNS. Furthermore, at least in the early stages of HE, neurones are likely exposed to less NH4+ than astrocytes as NH4+ is detoxified in astrocytes (Cooper et al. 1979). However, differences in NH4+ influx pathways may also contribute to the greater susceptibility of astrocytes to hyperammonaemia. Whereas NH4+ influx was primarily mediated by Na+, K+ ATPase in neurones, NH4+ influx into astrocytes was mediated by both Na+, K+ ATPase and NKCC activity. In addition to NH4+-evoked acidifications, hippocampal astrocytes exhibited large increases in [Na+]i caused by NKCC activity. These large [Na+]i increases reduce the inward driving force for Na+ ions, which many transport mechanisms rely upon. Previously, we showed that Na+-dependent glutamate uptake into cultured astrocytes was reduced in the presence of NH4+, an effect attributed to [Na+]i increases mediated by NKCC activity (Kelly et al. 2009). The rapid time course of events presented in this paper reflects the time course of events following rapid NH4+/NH3 increases in the CNS during acute HE. Animal models of acute HE, using bolus injection of NH4+/NH3, exhibit increases in extracellular glutamate within 20–40 min. Furthermore, behavioural deficits in animal models of acute HE also occurred on a rapid time course with coma and death occurring at 15–20 and 30–40 min, respectively (Hermenegildo et al. 2000). In contrast to acute HE, the progression of chronic HE is more prolonged and slower mechanisms play a role. However, in the present study we found that NH4+/NH3-evoked alterations in [Na+]i and pHi were maintained in the continued presence of NH4+/NH3 and therefore there is no a priori reason that NH4+/NH3-evoked alterations in [Na+]i and pHi do not also play a role in chronic HE.

 2010 The Authors Journal of Neurochemistry  2010 International Society for Neurochemistry, J. Neurochem. (2010) 115, 1123–1136

1134 | T. Kelly and C. R. Rose

In addition, [Na+]i increases caused by increases in expression and altered post-translational modifications of NKCC contribute to the swelling observed in cultured astrocytes 24 hrs following NH4+/NH3 application (Jayakumar et al. 2008). Other studies have also shown astrocytic swelling within minutes following NH4+/NH3 exposure (Alvarez-Leefmans et al. 2006; Reinehr et al. 2007). The rapid increases in [Na+]i observed in astrocytes in the present study provide a possible mechanism for the rapid cellular swelling (Dierkes et al. 2006; see Kelly et al. 2009). Taken together, the present study provides mechanisms by which NH4+ enters neurones and astrocytes in hippocampal slices, a prerequisite for the action of NH4+ on intracellular targets. The mechanisms of NH4+ influx differs between astrocytes and neurones; Na+, K+ ATPase primarily mediated NH4+ influx into neurones and both Na+, K+ ATPase and NKCC activity mediated NH4+ influx into astrocytes. In addition, influx of NH4+ was coupled to intracellular acidifications in both neurones and astrocytes, and additional [Na+]i increases in astrocytes. It should be noted that the study was performed on juvenile animals and that the relative contribution of the different mechanisms might change in the adult or aged brain. Nonetheless, our study indicates that differences in the influx pathways of NH4+ may contribute to the differential effects of NH4+ on hippocampal astrocytes and neurones during HE.

Acknowledgements This study was supported by Deutsche Forschungsgemeinschaft through Sonderforschungsbereich 575 ‘Experimental Hepatology’, TPC7, Du¨sseldorf. The authors declare no conflict of interest.

Supporting information Additional Supporting Information may be found in the online version of this article: Appendix S1. Supporting information. As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organised for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

References Ahboucha S. and Butterworth R. F. (2004) Pathophysiology of hepatic encephalopathy: a new look at GABA from the molecular standpoint Metab. Brain Dis. 19, 331–343. Alvarez-Leefmans F. J., Herrera-Perez J. J., Marquez M. S. and Blanco V. M. (2006) Simultaneous measurement of water volume and pH in single cells using BCECF and fluorescence imaging microscopy. Biophys. J. 90, 608–618. Bak L. K., Schousboe A. and Waagepetersen H. S. (2006) The glutamate/GABA-glutamine cycle: aspects of transport, neuro-

transmitter homeostasis and ammonia transfer. J. Neurochem. 98, 641–653. Berridge M. J. (1998) Neuronal calcium signaling. Neuron 21, 13–26. Bevensee M. O., Schwiening C. J. and Boron W. F. (1995) Use of BCECF and propidium iodide to assess membrane integrity of acutely isolated CA1 neurons from rat hippocampus. J. Neurosci. Methods 58, 61–75. Blanco G. and Mercer R. W. (1998) Isozymes of the Na-K-ATPase: heterogeneity in structure, diversity in function. Am. J. Physiol. 275, F633–F650. Boron W. F. (1983) Transport of H+ and of ionic weak acids and bases. J. Membr. Biol. 72, 1–16. Calabresi P., Marfia G. A., Centonze D., Pisani A. and Bernardi G. (1999) Sodium influx plays a major role in the membrane depolarization induced by oxygen and glucose deprivation in rat striatal spiny neurons. Stroke 30, 171–179. Casey J. R., Grinstein S. and Orlowski J. (2010) Sensors and regulators of intracellular pH. Nat. Rev. Mol. Cell Biol. 11, 50–61. Chatton J. Y., Marquet P. and Magistretti P. J. (2000) A quantitative analysis of L-glutamate-regulated Na+ dynamics in mouse cortical astrocytes: implications for cellular bioenergetics. Eur. J. Neurosci. 12, 3843–3853. Chesler M. (2003) Regulation and modulation of pH in the brain. Physiol. Rev. 84, 1183–1221. Cooper A. J., McDonald J. M., Gelbard A. S., Gledhill R. F. and Duffy T. E. (1979) The metabolic fate of 13N-labeled ammonia in rat brain. J. Biol. Chem. 254, 4982–4992. Deitmer J. M. and Rose C. R. (1996) pH regulation and proton signalling by glial cells. Prog. Neurobiol. 48, 73–103. Dierkes P. W., Wusten H. J., Klees G., Muller A. and Hochstrate P. (2006) Ionic mechanism of ouabain-induced swelling of leech Retzius neurons. Pflugers Arch. 452, 25–35. Ding D., Moskowitz S. I., Li R., Lee S. B., Esteban M., Tomaselli K., Chan J. and Bergold P. J. (2000) Acidosis induces necrosis and apoptosis of cultured hippocampal neurons. Exp. Neurol. 162, 1–12. Dobretsov M. and Stimers J. R. (2005) Neuronal function and alpha3 isoform of the Na/K-ATPase. Front. Biosci. 10, 2373– 2396. Felipo V. and Butterworth R. F. (2002) Neurobiology of ammonia. Prog. Neurobiol. 67, 259–279. Fiacco T. A. and McCarthy K. D. (2006) Astrocyte calcium elevations: properties, propagation, and effects on brain signaling. Glia 54, 676–690. Go¨rg B., Morwinsky A., Keitel V., Qvartskhava N., Schro¨r K. and Ha¨ussinger D. (2010) Ammonia triggers exocytotic release of L-glutamate from cultured rat astrocytes. Glia 58, 691–705. Hermenegildo C., Monfort P. and Felipo V. (2000) Activation of N-methyl-D-aspartate receptors in rat brain in vivo following acute ammonia intoxication: characterization by in vivo brain microdialysis. Hepatology 31, 709–715. Hertz L. and Kala G. (2007) Energy metabolism in brain cells: effects of elevated ammonia concentrations. Metab. Brain Dis. 22, 199– 218. Hindfelt B. (1975) L-Methionine DL-sulphoximine (MSO) and ammonia distribution between extra- and intracellular compartments of the rat brain. J. Neurol. Sci. 25, 499–506. Holm L. M., Jahn T. P., Møller A. L., Schjoerring J. K., Ferri D., Klaerke D. A. and Zeuthen T. (2005) NH3 and NH4+ permeability in aquaporin-expressing Xenopus oocytes. Pflugers Arch. 450, 415–428. Huang C. H. and Ye M. (2010) The Rh protein family: gene evolution, membrane biology, and disease association. Cell. Mol. Life Sci. 67, 1203–1218.

 2010 The Authors Journal of Neurochemistry  2010 International Society for Neurochemistry, J. Neurochem. (2010) 115, 1123–1136

NH4+/NH3 influx into neurones and astrocytes | 1135

Jayakumar A. R., Rama Rao K. V., Schousboe A. and Norenberg M. D. (2004) Glutamine-induced free radical production in cultured astrocytes. Glia 46, 296–301. Jayakumar A. R., Liu M., Moriyama M., Ramakrishnan R., Forbush B., Reddy P. V. B. and Norenberg M. D. (2008) Na–K–Cl cotransporter-1 in the mechanism of ammonia-induced astrocyte swelling. J. Biol. Chem. 283, 33874–33882. Kafitz K. W., Meier S. D., Stephan J. and Rose C. R. (2008) Developmental profile and properties of sulforhodamine 101-labeled glial cells in acute brain slices of rat hippocampus. J. Neurosci. Methods 169, 84–92. Kelly T., Kafitz K. W., Roderigo C. and Rose C. R. (2009) Ammoniumevoked alterations in intracellular sodium and pH reduce glial glutamate transport activity. Glia 57, 921–934. Knepper M. A., Packer R. and Good D. W. (1989) Ammonium transport in the kidney. Physiol. Rev. 69, 179–249. Langer J. and Rose C. R. (2009) Synaptically induced sodium signals in hippocampal astrocytes in situ. J. Physiol. 587, 5859– 5877. Liu X., Titz S., Lewen A. and Misgeld U. (2003) KCC2 mediates NH4+ uptake in cultured rat brain neurons. J. Neurophysiol. 90, 2785– 2790. Lockwood A. L. (2004) Blood ammonia levels and hepatic encephalopathy. Metab. Brain Dis. 19, 345–349. Lockwood A. L., Finn R. D., Campbell J. A. and Richman T. B. (1980) Factors that affect the uptake of ammonia by the brain: the blood– brain pH gradient. Brain Res. 181, 259–266. Lu J., Karadsheh M. and Delpire E. (1999) Developmental regulation of the neuronal-specific isoform of K–Cl cotransporter KCC2 in postnatal rat brains. J. Neurobiol. 39, 558–568. Marcaggi P. and Coles J. A. (2001) Ammonium in nervous tissue: transport across cell membranes, fluxes from neurons to glial cells, and role in signalling. Prog. Neurobiol. 64, 157–183. Marcaggi P., Thwaites D. T., Deitmer J. W. and Coles J. A. (1999) Chloride-dependent transport of NH4+ into bee retinal glial cells. Eur. J. Neurosci. 11, 167–177. Martinez A. (1968) Electron microscopy in human hepatic encephalopathy. Acta Neuropathol. 11, 82–86. McCarren M. and Alger B. E. (1987) Sodium-potassium pump inhibitors increase neuronal excitability in the rat hippocampal slice: role of a Ca2+-dependent conductance. J. Neurophysiol. 57, 496–509. McGrail K. M., Phillips J. M. and Sweadner K. J. (1991) Immunofluorescent localization of three Na,K-ATPase isozymes in the rat central nervous system: both neurons and glia can express more than one Na,K-ATPase. J. Neurosci. 11, 381–391. Meier S. D., Kovalchuk Y. and Rose C. R. (2006) Properties of the new fluorescent Na+ indicator CoroNa Green: comparison with SBFI and confocal Na+ imaging. J. Neurosci. Methods 155, 251– 259. Monfort P., Mun˜oz M. D., ElAyadi A., Kosenko E. and Felipo V. (2002) Effects of hyperammonemia and liver failure on glutamatergic neurotransmission. Metab. Brain Dis. 17, 237–250. Monteith G. R. and Blaustein M. P. (1998) Different effects of low and high dose cardiotonic steroids on cytosolic calcium in spontaneously active hippocampal neurons and in co-cultured glia. Brain Res. 795, 325–340. Moser H. (1987) Electrophysiological evidence for ammonium as a substitute for potassium in activating the sodium pump in a crayfish sensory neuron. Can. J. Physiol. Pharmacol. 65, 141–145. Nagaraja T. N. and Brookes N. (1998) Intracellular acidification induced by passive and active transport of ammonium ions in astrocytes. Am. J. Physiol. Cell Physiol. 274, C883–C891.

Norenberg M. D. (1977) A light and electron microscopic study of experimental portal-systemic (ammonia) encephalopathy. Progression and reversal of the disorder. Lab. Invest. 36, 618– 627. Norenberg M. D. and Lapham L. W. (1974) The astrocyte response in experimental portal-systemic encephalopathy: an electron microscopic study. J. Neuropathol. Exp. Neurol. 33, 422–435. Norenberg M. D., Rama Rao K. V. and Jayakumar A. R. (2004) Ammonia neurotoxicity and the mitochondrial permeability transition. J. Bioenerg. Biomembr. 36, 303–307. Norenberg M. D., Jayakumar A. R., Rama Rao K. V. and Panickar K. S. (2007) New concepts in the mechanism of ammonia-induced astrocyte swelling. Metab. Brain Dis. 22, 219–234. O’Brien W. J., Lingrel J. B. and Wallick E. T. (1994) Ouabain binding kinetics of the rat alpha two and alpha three isoforms of the sodium-potassium adenosine triphosphate. Arch. Biochem. Biophys. 310, 32–39. Payne J. A. (1997) Functional characterization of the neuronal-specific K–Cl cotransporter: implications for [K+]o regulation. Am. J. Physiol. Cell Physiol. 273, C1516–C1525. Pisani A., Calabresi P., Tozzi A., Bernardi G. and Knopfel T. (1998) Early sodium elevations induced by combined oxygen and glucose deprivation in pyramidal cortical neurons. Eur. J. Neurosci. 10, 3572–3574. Reinehr R., Go¨rg B., Becker S., Qvartskhava N., Bidmon H. J., Selbach O., Haas H. L., Schliess F. and Ha¨ussinger D. (2007) Hypoosmotic swelling and ammonia increase oxidative stress by NADPH oxidase in cultured astrocytes and vital brain slices. Glia 55, 758– 771. Roos A. and Boron W. F. (1981) Intracellular pH. Physiol. Rev. 61, 296– 434. Rose C. (2002) Increased extracellular brain glutamate in acute liver failure: decreased uptake or increased release? Metab. Brain Dis. 17, 251–261. Rose C. R. and Ransom B. R. (1996a) Intracellular sodium homeostasis in rat hippocampal astrocytes. J. Physiol. 491, 291–305. Rose C. R. and Ransom B. R. (1997) Regulation of intracellular sodium in cultured rat hippocampal neurones. J. Physiol. 499, 573–587. Rose C., Kresse W. and Kettenmann H. (2005) Acute insult of ammonia leads to calcium-dependent glutamate release from cultured astrocytes, an effect of pH. J. Biol. Chem. 280, 20937–20944. Russell J. M. (2000) Sodium–potassium–chloride cotransport. Physiol. Rev. 80, 211–276. Schneider D., Gerhardt E., Bock J., Muller M. M., Wolburg H., Lang F. and Schulz J. B. (2004) Intracellular acidification by inhibition of the Na+/H+-exchanger leads to caspase-independent death of cerebellar granule neurons resembling paraptosis. Cell Death Differ. 11, 760–770. Sheldon C., Diarra A., Cheng Y. M. and Church J. (2004) Sodium influx pathways during and after anoxia in rat hippocampal neurons. J. Neurosci. 24, 11057–11069. Skou J. C. (1960) Further investigations on a Mg++ + Na+-activated adenosintriphosphatase, possibly related to the active, linked transport of Na+ and K+ across the nerve membrane. Biochim. Biophys. Acta 42, 6–23. Stosiek C., Garaschuk O., Holthoff K. and Konnerth A. (2003) In vivo two-photon calcium imaging of neuronal networks. Proc. Natl Acad. Sci. USA 100, 7319–7324. Takeuchi A., Tatsumi S., Sarai N., Terashima K., Matsuoka S. and Noma A. (2006) Ionic mechanisms of cardiac cell swelling induced by blocking Na+/K+ pump as revealed by experiments and simulation. J. Gen. Physiol. 128, 495–507.

 2010 The Authors Journal of Neurochemistry  2010 International Society for Neurochemistry, J. Neurochem. (2010) 115, 1123–1136

1136 | T. Kelly and C. R. Rose

Thomas R. C. (1984) Review lecture: experimental displacement of intracellular pH and the mechanism of its subsequent recovery. J. Physiol. 354, 3P–22P. Thomas R. C. (2002) The effects of HCl and CaCl2 injections on intracellular calcium and pH in voltage-clamped snail (Helix aspersa) neurons. J. Gen. Physiol. 120, 567–579. Titz S., Hormuzdi S., Lewen A., Monyer H. and Misgeld U. (2006) Intracellular acidification in neurons induced by ammonium depends on KCC2 function. Eur. J. Neurosci. 23, 454–464. Towle D. W. (1984) Membrane-bound ATPases in arthropod ion-transporting tissues. Am. Zool. 24, 177–185. Vaquero J. and Butterworth R. F. (2007) Mechanisms of brain edema in acute liver failure and impact of novel therapeutic interventions. Neurol. Res. 29, 683–690. Wall S. M. (1997) Ouabain reduces net acid secretion and increases pHi by inhibiting NH4+ uptake on rat tIMCD Na(+)-K(+)-ATPase. Am. J. Physiol. 273, F857–F868.

Walz W. and Wuttke W. A. (1999) Independent mechanisms of potassium clearance by astrocytes in gliotic tissue. J. Neurosci. Res. 56, 595–603. Weissenborn K., Bokemeyer M., Krause J., Ennen J. and Ahl B. (2005) Neurological and neuropsychiatric syndromes associated with liver disease. AIDS 19, s93–s98. Yan Y., Dempsey R. J. and Sun D. (2001) Expression of Na+–K+–Cl) cotransporter in rat brain during development and its localization in mature astrocytes. Brain Res. 911, 43–55. Ytrebo L. M., Kristiansen R. G., Maehre H., Fuskevag O. M., Kalstad T., Revhaug A., Cobos M. J., Jalan R. and Rose C. F. (2009) L-Ornithine phenylacetate attenuates increased arterial and extracellular brain ammonia and prevents intracranial hypertension in pigs with acute liver failure. Hepatology 50, 165–174. Zhang Y. and Barres B. A. (2010) Astrocyte heterogeneity: an underappreciated topic in neurobiology. Curr. Opin. Neurobiol. 20, 588– 594.

 2010 The Authors Journal of Neurochemistry  2010 International Society for Neurochemistry, J. Neurochem. (2010) 115, 1123–1136