Brain Research, 541 (1991) 103-109
103
Elsevier
BRES 16314
Activation of neurokinin receptors modulates K + and C1- channel activity in cultured astrocytes from rat cortex Kurt Harald Backus*, Thomas Berger and Helmut Kettenmann Department of Neurobiology, University of Heidelberg, Heidelberg (ER. G.) (Accepted 4 September 1990)
Key words: Glia; Substance P; Ion channel; Potassium homeostasis; Electrophysioloy
Short application of the neurokinin receptor agonist substance P (SP) leads to a biphasic depolarization of astrocytes cultured from rat cortex. The rapid and transient depolarizing event lasted few seconds, the slow one several minutes. In some cells, only the slow depolarizing component was observed. During the slow depolarizing event, the sensitivity of the membrane potential for a change in the K + gradient decreased, indicating a decrease in the relative K + permeability of the membrane. The rapid SP-induced depolarization could be reversed, when the membrane potential was depolarized to about 0 mV by elevation of the extracellular K + concentration, indicating a reversal potential close to the Cl- equilibrium potential. When the membrane was damped close to the resting membrane potential using the whole-ceU patch-clamp technique, SP induced a biphasic inward current with a similar time course as the SP-induced membrane depolarization. Evaluating current-to-voltage curves indicated a conductance decrease during the slow inward current with a reversal potential of the SP-dependent current close to the K ÷ equilibrium potential. The mean open time of single K + channels, measured in the cell-attached configuration of the patch-clamp technique, decreased after application of SP. In contrast, the mean open time of single Cl- channels increased. We conclude that activation of neurokinin receptors in astrocytes modulates the activity of K + and CI- channels, leading to a complex depolarization of the membrane potential. INTRODUCTION In recent years, several n e u r o t r a n s m i t t e r receptors have been identified on glial cells 17. These receptors include those which are directly linked to ionic channels such as the G A B A and glutamate r e c e p t o r of astrocytes 6' 14,34 as well as those which are most likely linked to K ÷ channels via second messengers or G proteins such as the a - a d r e n e r g i c r e c e p t o r 11. In addition to the 'classical' n e u r o t r a n s m i t t e r receptors, glial cells also express receptors for n e u r o p e p t i d e s , e.g. receptors binding neurokinin agonists 5'35 and bradykinin 28. Recently, we could demonstrate that activation of neurokinin receptors leads to changes in the m e m b r a n e potential of cultured astrocytes from rat cortex 37. Brief application of neurokinin agonists, such as SP or neurokinin A , led to a rapid and transient d e p o l a r i z a t i o n followed by a long-lasting (several minutes) depolarization of the m e m b r a n e . The ionic mechanism underlying this depolarization was unidentified. In neurons and muscle cells, activation of SP receptors affects a variety of different m e m b r a n e channels. In bullfrog sympathetic neurons and toad smooth muscle cells, SP and acetylcholine inhibit M currents 1' 12,30. A c t i v a t i o n of SP receptors also exerts a direct effect
on acetylcholine-induced currents 8. In spinal dorsal horn cells of the rat, SP augments a calcium-sensitive current 21, while in guinea pig inferior mesenteric ganglion cells, b o t h decreases and increases in m e m b r a n e conductance were observed, which were not m e d i a t e d by M currents 9. A closure of K ÷ channels by activation of SP receptors was o b s e r v e d in spinal cord n e u r o n s 26 which is a c c o m p a n i e d by an increase in the N a ÷ p e r m e a b i l i t y in inferior mesenteric ganglion cells of the guinea pig 19. These results illustrate that n e u r o k i n i n r e c e p t o r s can be coupled to a variety of different ion channels, and that the set of channels influenced by neurokinins varies among different cell types. T h e link b e t w e e n SP r e c e p t o r activation and the conductance change was identified in rat pancreatic acinar ceils as an IP3-1inked calcium mobilization 31. In neurons from nucleus basalis a pertussis toxin-insensitive G p r o t e i n m e d i a t e d SP-induced inhibition of inward rectifying K ÷ channels 23. Cultured astrocytes express a variety o f m e m b r a n e channels including different types of K ÷ channels 4'2s'33, N a ÷ channels 3 and CI- channels 2'25. These channels could be involved in the SP-induced m e m b r a n e depolarization. The aim of the present study was t h e r e f o r e to identify the ionic mechanism underlying the SP-induced m e m b r a n e
* Present address: Pharmaceutical Research Department, E Hoffmann-La Roche, Grenzacherstrasse 124, CH-4002 Basel, Switzerland
Correspondence: H. Kettenmann, Department of Neurobiology, University of Heidelberg, Im Neuenheimer Feld 345, 6900 Heidelberg, F.R.G. 0006-8993/91/$03.50 © 1991 Elsevier Science Publishers B.V. (Biomedical Division)
104 depolarization in cultured astrocytes of the rat. We demonstrate that activation of neurokinin receptors led to a closure of K ÷ channels and the opening of C1channels.
MATERIALS AND METHODS
Cell culture Cultures of enriched astrocytes ( > 90%) were obtained from cerebral hemispheres of 0- to 2-day-old Sprague-Dawley rats, according to McCarthy and de Vellis TM with modifications described by Keilhauer et alJ 3, and maintained in culture for 3-6 weeks. We stained sister cultures to those used in the electrophysioiogical experiments with GFAP antibodies to label the astrocytes in that culture system. These cells could be easily distinguished from the contaminating macrophages/microglial cells based on morphological criteria.
Solutions Cultures were continuously superfused during the experiments. Within 1-3 s 90% of the bathing fluid could be exchanged. The standard salt solution for intracellular recordings contained (in mM): KCI 5.4, NaC1 116, NaH2PO4.H20 1.0, MgSO 4 0.8, CaCI 2 1.8, D-glucose 5.6, NaHCO 3 26.2, HEPES 10. The pH was adjusted to 7.4. In some experiments NaCi was equimolarly replaced by KC! to depolarize the cells by increasing extracellular K ÷ concentration ([K+]o) from 5.4 to 50 or 121 mM. To alter both the K ÷ and CIgradient, NaC1 was replaced by potassium acetate. For patch-clamp recordings, the standard salt solution contained (in mM): KCI 5.4, NaCI 116, NaH2POa.H20 1.0, MgCI 2 0.8, CaCI 2 1.8, D-glucose 11.1, NaHCO 3 26.2, HEPES 5. The pH was adjusted to 7.2. Substance P (10-5"M; Sigma, Munich, E R . G ) was added to the standard salt solutions and applied by bath perfusion.
lmracellular recording Cultures were maintained on the stage of an inverted microscope at about 25 °C. For recording the membrane potential, cells were penetrated by 2 M KCl-filled electrodes (resistance 40-80 MI2) with the aid of a step-motor driven micromanipulator 32. Data were amplified by a WPI 707 amplifier and transferred to a computer system.
Patch-clamp procedures Whole-cell and single-channel currents were recorded with an EPC-7 amplifier (List, Darmstadt, F.R.G.) using standard methods of the patch-clamp technique ~°. Pipettes for whole cell recordings contained (in raM): KCI 100, NaCI 30, MgCI 2 1, CaCI 2 0.5, EGTA 5, HEPES 10, ATP 3, GTP 0.1. For single-channel recording, pipettes contained: KCI 146, MgCI 2 1.2, CaC12 2.6, HEPES 10. The pH was adjusted to 7.2. To determine the reversal potential of the SP response, currents were measured in the whole-cell-clamp configuration during application of a voltage ramp. The voltage ramp ranged from -150 to 90 mV and lasted 1.6 s.
Data processing Data were digitized by an interface card connected to an AT-compatible computer. Data for single-channel analysis were filtered at 2 kHz (Bessel 8-pole) and sampled at a frequency of 5 kHz. A Pascal program package developed in our laboratory was used to display and analyze the recorded traces. The program contained routines to determine current amplitudes and open state probabilities of single-channel currents. In addition, the computer was connected to the voltage control of the patch-clamp amplifier and generated the voltage changes during recording.
RESULTS
SP-induced membrane depolarizations Application of SP for 10 s induced a depolarization of the membrane potential in cultured astrocytes from rat cortex (Fig. 1, inset). The mean depolarization was 4.5 mV (n = 6) and lasted for minutes as has been previously described 37. The depolarization was preceded in about half of the experiments by a rapid, transient depolarization of about 4 mV, lasting few seconds.To characterize the slow depolarization, we analyzed the K ÷ sensitivity of the membrane by altering the K ÷ gradient. When [K+]o was increased from 5.4 to 50 mM, the membrane
Vm [mV]
1~SP
-30 -40
. I0 mV
-50 -60 -7o
q 'l~sP 5
min
Fig. 1. Sensitivity of the membrane potential for a change in the K ÷ gradient after application of SP. During recording of membrane potential (I'm) from an astrocyte with a conventional microelectrode, extracellular K + concentration was repeatedly elevated from 5.4 mM to 50 mM for 45 s as denoted by bars. SP was applied for 10 s as indicated by arrow. The inset on top shows the time course of an SP-induced membrane depolarization without altering the K + gradient. Bars in the inset denote 10 mV and 1 min.
121 m M K +
Fig. 2. Reversal of the first transient depolarization. During recording of membrane potential (Vm) from an astrocyte with a conventional microelectrode, SP was applied for 10 s as indicated by arrow. Resting membrane potential was -70 mV. A magnification of the SP-induced response is shown in the inset. After 25 rain the membrane was depolarized by elevating [K+]o from 5.4 to 121 mM. When SP (arrow) was again applied under these depolarized conditions, a transient hyperpolarizing response was elicited. The inset on the right displays a magnified trace.
105
®
®
(~
[nil
contro.~
TITI II1=I11I-I it 1 II II I II/]'TI'I
I
SP
1 min
o
sP
I I
gt/go
I SP I
1.1 o
1
• ~ D
OD
E
ol
÷
o I
OOOiD
• Io
oo
D QOO
oo
0.8 0.7 Fig. 3. Changes in membrane conductance induced by SP. A: membrane currents from an astrocyte with a large and flattened morphological appearance were recorded using the patch-clamp technique in the whole-cell recording configuration. The membrane potential was clamped at --60 mV and was repetitively switched for 400 ms intervals to -90 mV. SP was applied as indicated by a dashed line. The relative conductance (conductance at the indicated time (g0/conductance at the beginning of the recording (go)) is displayed corresponding to the current on top. B: a similar recording as described in the legend to 3A was obtained from a smaller cell (soma diameter less than 10/~m). SP did not elicit a detectable response. The top trace displays the current recording, the bottom trace the relative input conductance.
potential of astrocytes depolarized by about 50 mV, as previously described 2°. The size of the depolarization reflects the relative K ÷ conductance. After application of SP, K÷-induced depolarizations decreased in size, indicating a decrease of the relative K ÷ conductance (Fig. 1; n = 5). We analyzed the fast depolarizing event with an experimental paradigm previously applied to characterize G A B A - a c t i v a t e d CI- channels in astrocytes 15. SP was, therefore, applied under normal conditions and after depolarizing the membrane close to 0 mV by increasing [K+]o to 121 m M (Fig. 2). In elevated [K+]o, SP induced a hyperpolarizing response (1 mV, S.D. 0.6, n = 7) with a time course similar to the fast depolarization seen in normal [K+]o . To shift the CI- gradient to a more positive value under these depolarized conditions, 116 mM NaCI was replaced by an equimolar amount of potassium acetate ([Cl-]o was at 10 raM). The membrane potential again depolarized close to 0 mV in high [K+]o, but no response could be observed after administration of SP (not shown; n = 4). These experiments suggest an increase of a CI- conductance during the rapid depolarizing event. SP-induced currents in cultured astrocytes
We analyzed whole cell currents with the patch-clamp technique to determine the changes in conductance induced by SP, when the membrane was clamped close to
0_J___ _ ~.............. l
-120
Vm -60
t~ [mV]
Fig. 4. Reversal potential of the slow SP response. A: currentvoltage (l-V) curves were obtained by applying a voltage ramp changing the membrane potential (Vrn) from -120 to 0 mV while recording current (/) in the whole-cell recording mode of the patch-clamp technique. The I-V curve prior to application of SP (control) and 3 min after application of SP at the peak of the slow inward current (SP) are shown. B: the I - V curve of the SP-sensitive current was obtained by subtracting the 1-V curve after application of SP from that of the control (SP-control). The potassium equilibrium potential (EK+) is indicated by the dashed vertical line.
the resting potential. In contrast to the previous study with conventional electrodes aT, these recordings resolved two populations of cells with respect to SP sensitivity: about half of the cells responded with a change in the resting current to application of SP (Fig. 3A), while another population was unresponsive (Fig. 3B). These unresponsive cells (n = 22) had small somata ( < 10/tm) and were further characterized by a high input resistance (R i = 848 Mg2; range 40-4300 MI2). These cells were too small to be successfully impaled with conventional electrodes. In contrast, responsive cells (n = 23) had a large and flattened morphological appearance, a much lower input resistance R i = 168 Mg2 (range 5-1090 MI2; n = 16) and could be impaled with conventional microelectrodes. In some of the responsive cells (n = 7), the SP-induced current had a monophasic time course with a mean peak of 250 p A (range 5-1000 pA). In all other responsive cells (n = 16) the inward current was biphasic; a fast transient inward current (peak at 438 p A ; range 100-1500 pA) that inactivated within 30-60 s was followed by a slow and longer lasting inward current. The latter current peaked after about 2-5 min with an average amplitude of 534 p A (range 30-2800 pA). It slowly returned, but commonly did not reach the previous level even after several minutes. Both types of cells were GFAP-positive when sister cultures were immunolabelled.
106
-30 I[pA]
t I, V _ I [ r n V ] ~ " " -60
,
i
5e 100 ms
SP+7 min
-~6
-
=1 IS . 10(1 m s
®
® PO
4 0 ~
[*/*l 20
0
10
20 rain
Fig. 5. Recordings from K÷ channels. A: a current (/) to voltage (V) curve was obtained from single channel recordings with the patch-clamp technique in the cell-attached configuration. The insets show two sample recordings at -18 and -78 mV (calibration bars refer to the inset traces). The K+ (EK+) and CI- (EcF) equilibrium potentials are indicated by the dashed lines. The extrapolated reversal potential of the current is close to EK+ identifying this channel as a K + channel. B: this scheme illustrates the electrode arrangement in the cell attached recording configuration. SP was applied to the part of the membrane which was not covered by the pipette. C: sample traces of current recordings from the K ÷ channels shown in A prior to SP application (control) and 7 (SP+7 min) and 23 min (SP+23 min) after SP application. The pipette contained 146 mM K ÷ and 154 mM CF. D: from intact patch recordings of 26 s duration, the mean open state probability of the first current level (Po) was determined before and at different times after a 10 s SP application (indicated by the dashed line).
Effect of SP on cell input conductance To determine the conductance changes during an SP response, we recorded currents at two different membrane potentials by briefly changing the holding potential (-60 mV) to a more negative value (-90 mV). These potential jumps induced inward currents. After applying SP, the amplitude of these currents decreased in most cells (Fig. 3A). This finding indicates that the slow component of the SP response is accompanied by a
I [pA]
(~
control
[1 pA 100 ms
SP + 18 rain
I V
-150 -100 -5{2
50
[mV]
-2
®
conductance decrease. It was not possible to resolve conductance changes during the fast, transient response and all subsequent data were obtained at the peak of the slow, sustained phase. To further characterize these conductance changes, we determined current-to-voltage curves of the membrane before and after application of SP. The membrane potential of astrocytes was continuously changed between -150 and 90 mV within 1.6 s by a voltage-ramp to determine steady-state current-voltage (I-V) relations. I - V curves were recorded before (control) and 2 min after application of SP (10-5 M) (at the peak of the slow, sustained response), and the slopes representing cell input conductance were compared. During the application of SP the input conductance decreased during the slow, sustained phase in a similar way as has been observed from recordings with rectangular current pulses (Fig. 4). From these two approaches we determined the mean input conductance changes. During the slow, sustained phase the input conductance showed a decrease to 95% (S.D. = 5%; n = 15) indicating the closure of ion channels during that phase. In many cases we observed that the increase in input resistance was not fully reversible and continued to last for the rest of the recording time (up to more than one hour). Under these conditions, a second application of SP induced only a very small or even no detectable current response.
Reversal potential of the SP-sensitive current The reversal potential of the SP-dependent current was determined to characterize the ionic species mediating the SP-induced changes. We could only analyze the slow component of the SP effect, since the inactivation of the fast, transient component was too fast to obtain the I - V relation. I - V curves at the peak of the slow inward current response were subtracted from those obtained under control conditions. This subtraction yielded the current component which was blocked after application of SP (Fig. 4). The mean reversal potential of this current was -93 mV (S.D. 25 mV, n = 5) and thus close to the K ÷ equilibrium potential (EK = -81 mV) indicating that K ÷ channels close during the slow, sustained phase.
Possible effect of SP on M-type K + channels
® [v.]
40
30 20 10 0
10
20 min
Fig. 6. Single channel recordings from a CI- channel. A. CI -selective channel was identified by its reversal potential close to the Cl--equilibrium potential. For details see legend to Fig. 5.
Several groups have reported that the action of neurokinins on neuronal membranes is mediated via a blockage of M-type K + channels 1'26. We applied the voltage-clamp protocol designed to isolate M currents 1 before and during application of SP. The membrane was clamped at depolarized membrane potentials (-15 or -35 mV) and hyperpolarizing voltage steps close to the resting potential (-60 mV) were applied for 0.4-2 s.
107
Fig. 7. Scheme of neurokinin receptor involvement in [K+]o homeostasis. K÷ is released during electrical activity of a neuron (N; series of action potentials displayed in the inset). Astrocytes (A) take up and redistribute the excess K+ from the extraceUular space. Local activation of glial neurokinin receptors alter the direction of the K÷ currents. For details see discussion.
Under these conditions, no relaxation of an outward current was observed during or after application of SP (not shown; n = 6). We therefore conclude that astrocytes under our culture conditions did not express M-type K ÷ channels which are affected by SP.
Identification of K ÷ and CI- channels Single channel recordings were used to further characterize the interaction between neurokinin receptor activation and its effect on electrical membrane properties. Before or after testing the effect of SP on the single-channel properties, the reversal potential was determined to identify the ionic species permeating the channel. To compare reversal potentials with the equilibrium potentials for K + and Cl-, we made the following assumptions: the resting membrane potential was assumed at -63 mV (mean value determined at the beginning of whole-cell recordings; S.D. 5 mV, n = 26), intracellular [K +] at 100 mM 36, intracellular [Cl-] at 40 mM 15. When using 146 mM K + and 154 mM Cl- in the pipette for cell-attached recordings, the K + equilibrium potential was at 10 mV, the Cl- equilibrium potential at -34 mV. Voltage ramps ranging from 57 mV to -183 mV (membrane potential values based on a mean resting potential as described above) were applied to determine the reversal potentials of single-channel currents (not shown). In other experiments, we determined the reversal potential from continuous recordings at different membrane potentials (see Figs. 5A and 6A). We identified two populations of channels, one population with a reversal potential close to the K + equilibrium potential (n = 29), and one with a reversal potential close to the Clequilibrium potential (n = 5).
Effect of SP on single-channel properties While recording from single channels in the cellattached configuration of the patch-clamp technique (Fig. 5B), neurokinin receptors were activated by applying SP (10-5 M) to the bath for 10 s. We analyzed mean open time and the decay time constant of open and closed time histograms of the single channel properties before, during and after application of SP. In 8 out of 10 experiments, the mean open time of K ÷ channels determined from recordings of 26 s duration, decreased to 12% (S.D. 9.5%) of the control value within the first 7 min (Fig. 5C,D). The peak of the decrease in mean open time occurred after less than 2 min in 5 patches and after 6-7 min in the other 3 recordings. In 2 recordings we observed the opposite, namely an increase of the mean open time to 250% and 2300% of the control value. We selected 3 recordings which showed a decrease in open probability after application of SP and analyzed the decay constants of the open and closed times. These patches did not show any superposition of simultaneous openings. In these samples, the decay constant of the open times decreased to 64% of the control value, while the two decay constants of the closed times increased to 108% and to 250% of the control, respectively. The amplitude of the single-channel current was slightly reduced after application of SP; this was most likely due to the SP-induced depolarization.We were able to test the effect of SP on the mean open time of 2 out of 5 identified CIchannels (Fig. 6C,D). In both cases, mean open probability increased 9- and 15-fold, respectively, within 3 min. This increase was in part reversible. DISCUSSION
Neurokinin receptor activation and ion channel link In this study we provide evidence that the slow membrane depolarization in astrocytes is caused by an increased tendency of K ÷ channels to close. This can be deduced from the finding that the relative K + permeability during the depolarized phase decreased in concert with a decrease in membrane conductance. This observed conductance decrease is small, but it should be taken into account that we recorded from a large syncytium and that a large proportion of the injected current did spread to other cells via gap junctions which are numerously present in these cells 16. Additional evidence comes from recordings of single K ÷ channels. We observed that K ÷ channels were decreased in their open state probability after neurokinin receptor activation. This decrease is not due to membrane depolarization per se, since the mean open time of most K + channels in astrocytes increased with membrane depolarization. A more detailed characterization of the different types of K + channels in
108 astrocytes, however, is required to link the activation of neurokinin receptors to defined populations of K ÷ channels. The first transient depolarization which was observed in many, but not all astrocytes could be reversed by depolarizing the membrane potential in high K ÷ similar as has been shown for the G A B A activated CI- channel in astrocytes 15. This observation together with the notion that CI- channel open probability was increased after application of SP indicates that SP activates a C1conductance with a different time course as the decrease in K ÷ channel conductance. Both conductance changes are superimposed after activation of SP receptors and the different time courses could explain the complicated time courses of membrane depolarization 37 or inward currents. When recording from channels in the intact patch configuration, bath application of SP did not directly affect the channel under the pipette. We can therefore exclude a direct effect of SP on K ÷ or C1- channels and it is most likely that the link between SP receptor activation and channel modulation is mediated by second messengers e.g. by IP3-1inked calcium mobilization as described in rat pancreatic acinar cells31.
Comparison of glial and neuronal SP responses In neurons, brief application of neurokinin agonists induced a depolarization of the membrane potential similar to that described in this study for cultured astrocytes ~9. Neuronal neurokinin receptor activation can exert multiple and complex effects on the activity of different ion channels 9'22. This includes an inactivation of M currents 1 and an activation of a cation-permeability 29. In this study we did not observe changes in currents recorded according to an M-channel identification protocol after application of SP. The effect of neurokinin receptor activation on the CIconductance in astrocytes has so far not been described in any other cell system. It can therefore be concluded that neurokinin receptors are linked to specific astrocytic channels rather than that astrocytes express a neuronal-like type of neurokinin receptorqon channel system.
Heterogeneity of astrocytes with respect to SP responsiveness We found two populations of cells in the purified astrocytic cultures: large cells with a high input conductance which responded to SP. This population correREFERENCES 1 Adams, P.R., Brown, D.A. and Jones, S.W., Substance P inhibits the M-current in bullfrog sympathetic neurons, Br. J. Pharmacol., 79 (1983) 330-333. 2 Barres, B.A., Chun, L.L.Y. and Corey, D.P., Ion channel expression by white matter glia: I. Type 2 astrocytes and oligodendrocytes, Glia, 1 (1988) 10-30. 3 Bevan, S., Chiu, S.Y., Gray, P.T.A. and Ritchie, J.M., The
sponds to the electrically extensively coupled astrocytes which form a large syncytium in this culture system 16. The extent of electrical coupling is reflected in the high input conductance of these cells, a characteristic feature of the astrocytic syncytium. The smaller cells which were unresponsive to SP had a morphological appearance of more immature astrocytes which is reflected by their lower input conductance. It is likely that astrocytes express higher densities of SP receptors during maturation. This is also substantiated by our observations that SP responses increased in size with prolonged maintenance of cells in culture. Wilkin et al.38 also reported that SP receptor density increased with time in culture.
Functional implications of ion channel control by neurokinin receptors One of the important functions of astrocytes is the maintenance of the K + homeostasis in the extraceUular space. It was demonstrated recently that passive currents which redistribute K ÷ and thereby regulate [K+]o, can be facilitated by an uneven distribution of K + channels along glial cells 24. This allows a spatial control and enforces a direction of these currents. The modulation of K ÷ and CI- channels by neuropeptides could add a temporal component to this control. It can be speculated that in areas where SP receptors are activated and K ÷ channels are closed, K ÷ release and uptake are suppressed. Peptide receptors on astrocytes could therefore guide spatial buffer currents which could result in temporal changes in glial K ÷ regulation (Fig. 7). It has been demonstrated for Schwann cells that K ÷ channel activity is linked to the proliferative state of that ceil. Moreover, blockade of K ÷ channels led to a blockade of proliferation in the Schwann cells 7 and in retinal glial ceUs27. Assuming a similar link between proliferation and K ÷ channel activity in astrocytes, local levels of SP in the brain could regulate the proliferative state of these cells.
Acknowledgements. This research was supported by Bundesministerium fiir Forschung und Technologie and Deutsche Forschungsgemeinschaft (Heisenberg-Stipendium to H.K.I SFB 317). The authors thank Drs. H. Marrero, M. Schachner, J. Trotter, G. Trube and M. Wienrich for discussion and helpful comments on the manuscript, and T. Berger, S. Hauck, and S. Riese for excellent technical assistance. presence of voltage-gated sodium, potassium and chloride channels in rat cultured astrocytes, Proc. R. Soc. London, 225 (1985) 299-313. 4 Bevan, S. and Raft, M., Voltage-dependent potassium currents in cultured astrocytes, Nature, 315 (1985) 229-232. 5 Beaujouan, J.C., Daguet de Montety, M.D., Torrens, Y., Dietl, M. and Glowinski, J., Selective presence of a high density of NK-1 receptors on astrocytes from the rat brain stem in primary culture, Reg. Peptides, 22 (1988) 33.
109 6 Bormann, J. and Kettenmann, H., Patch clamp study of GABA receptor CI- channels in cultured astrocytes, Proc. Natl. Acad. Sci. U.S.A., 85 (1988) 9336-9340. 7 Chiu, S.Y. and Wilson, G.E, The role of potassium channels in Schwann cell proliferation in Wallerian degeneration of explant rabbit sciatic nerves, J. Physiol. (London), 408 (1989) 199-222. 8 Clapham, D.E. and Neher, E., Substance P reduces acetylcholine-induced currents in isolated bovine chromaffin cells, J. Physiol. (London), 347 (1984) 255-277. 9 Griffith, W.H., Hills, J.M. and Brown, D.A., Substance P-mediated membrane currents in voltage-clamped guinea pig inferior mesenteric ganglion cells, Synapse, 2 (1988) 432-441. 10 Hamill, O.P., Marty, A., Neher, E., Sakmann, B. and Sigworth F.J., Improved patch-clamp techniques for high-resolution current recording from cells and cell-free membrane patches, Pflagers Arch., 391 (1981) 85-100. 11 Hirata, H., Slater, N.T. and Kimelberg, H.K., Alpha-adrenergic receptor-mediated depolarization of rat neocortical astrocytes in primary culture, Brain Research, 270 (1983) 358-362. 12 Jones, S.W., Muscarinic and peptidergic excitation of bull-frog sympathetic neurones, J. Physiol. (London), 366 (1985) 63-87. 13 Keilhauer, G., Meier, D.H., Kuhlmann-Krieg, S., Nieke, J. and Schachner, M., Astrocytes support incomplete differentiation of an oligodendrocyte precursor cell, EMBO J., 44 (1985) 24992504. 14 Kettenmann, H., Backus, K.H. and Schachner, M., Aspartate, glutamate and gamma-aminobutyric acid depolarize cultured astrocytes, Neurosci. Lett., 52 (1984) 25-29. 15 Kettenmann, H., Backus, K.H. and Schachner, M., GABA opens Cl-channels in cultured astrocytes, Brain Research, 404 (1987) 1-9. 16 Kettenmann, H. and Ransom, B.R., Electrical coupling between astrocytes and between oligodendrocytes in mammalian cell cultures, Glia, 1 (1988) 64-73. 17 Kimelberg, H.K. (Ed.), Glial Cell Receptors, Raven Press, New York, 1988. 18 McCarthy, K.D. and De Vellis, J., Preparation of separate astroglial and oligodendroglial cell cultures from rat cerebral tissue, J. Cell Biol., 85 (1980) 890-902. 19 Minota, S., Dun, N.J. and Karczmar, A.G., Substance Pinduced depolarization in sympathetic neurons: not simple K-inactivation, Brain Research, 216 (1981) 224-228. 20 Moonen, G. and Nelson, P.G., Some properties of astrocytes in primary cultures. In E. Schoffeniels, G. Frank, L. Hertz and D.B.L. Tower (Eds.), Dynamic Properties of Glial Cells, Pergamon Press, London, 1978, pp. 389-393. 21 Murase, K., Ryu, P.D. and Randic, M., Substance P augments a persistent slow inward calcium-sensitive current in voltageclamped spinal dorsal horn neurons of the rat, Brain Research, 365 (1986) 369-376. 22 Murase, K., Ryu, P.D. and Randic, M., Tachykinins modulate multiple conductances in voltage-clamped rat spinal dorsal horn neurons, J. Neurophysiol., 61 (1989) 854-865. 23 Nakajima, Y., Nakajima, S. and Inoue, M., Pertussis toxin-
24 25 26 27 28
29 30
31
32
33 34 35
36
37 38
insensitive G protein mediates substance P-induced inhibition of potassium channels in brain neurons, Proc. Natl. Acad. Sci. U.S.A., 85 (1988) 3643-3647. Newman, E.A., Distribution of potassium conductance in mammalian Miiller (glial) cells: a comparative study, J. Neurosci., 7 (1987) 2423-2432. Nowak, L., Ascher, P. and Berwald-Netter, Y., Ionic channels in mouse astrocytes in culture, J. Neurosci., 7 (1987) 101-109. Nowak, L.M. and Macdonald, R.L., Substance P decreases a potassium conductance of spinal cord neurons in cell culture, Brain Research, 214 (1981) 416--423. Puro, D.G., Roberge, F. and Chan, C.-C., Retinal glial cells and ion channels: a possible link, lnvestigat. Ophthalmol. Visual Sci., 30 (1989) 521-529. Reiser, G. and Hamprecht, B., Bradykinin induces hyperpolarizations in rat glioma cells and in neuroblastoma x glioma hybrid cells, Brain Research, 239 (1982) 191-199. Reiser, G. and Hamprecht, B, Substance P and serotonin act synergistically to activate a cation permeability in a neuronal cell line, Brain Research, 479 (1989) 40-48. Sims, S.M., Walsh, J.V. and Singer, J.J., Substance P and acetylcholine both suppress the same K + current in dissociated smooth muscle cells, J. Physiol. (London), 251 (1986) C580-C587. Song, S.-Y., Iwashita, S., Noguchi, K. and Konishi, S., Inositol triphosphate-linked calcium mobilization couples substance P receptors to conductance increase in rat pancreatic acinar cell line, Neurosci. Lett., 95 (1988) 143-148. Sonnhof, U., Frrderer, R., Schneider, W. and Kettenmann, H., Cell puncturing with a step motor driven manipulator with simultaneous control of displacement, Pflagers Arch., 329 (1982) 295-300. Sonnhof, U. and Schachner, M., Single voltage-dependent K÷-channels in cultured astrocytes, Neurosci. Lett., 64 (1986) 241-246. Sontheimer, H., Kettenmann, H., Backus, K.H. and Schachner, M., Glutamate opens Na+/K + channels in cultured astrocytes, Glia, 1 (1988) 328-336. Torrens, Y., Beaujouan, J.C., Saffroy, M., Daguet de Montety, M.C., Bergstr6m, L. and Glowinski, J., Substance P receptors in primary cultures of cortical astrocytes from the mouse, Proc. Natl. Acad. Sci. U.S.A., 83 (1986) 9216-9220. Walz, W., Wuttke, W. and Hertz, L., Astrocytes in primary cultures: membrane potential characteristics reveal exclusive potassium conductance and potassium accumulator properties, Brain Research, 292 (1984) 367-374. Wienrich, M. and Kettenmann, H., Activation of substance P receptors leads to membrane potential responses in cultured astrocytes, Glia, 2 (1989) 155-160. Wilkin, G.P., Cholewinski, A.J., Reid, J.C. and Marriott, D.R., Peptide receptors on astrocytes: characterization and regional heterogeneity, ln: Differentiation and Functions of Glial Cells, Satellite Symposium of the 12th Biennal Meeting of the International Society for Neurochemistry, Rome, 1989, p. 36.
103
Elsevier
BRES 16314
Activation of neurokinin receptors modulates K + and C1- channel activity in cultured astrocytes from rat cortex Kurt Harald Backus*, Thomas Berger and Helmut Kettenmann Department of Neurobiology, University of Heidelberg, Heidelberg (ER. G.) (Accepted 4 September 1990)
Key words: Glia; Substance P; Ion channel; Potassium homeostasis; Electrophysioloy
Short application of the neurokinin receptor agonist substance P (SP) leads to a biphasic depolarization of astrocytes cultured from rat cortex. The rapid and transient depolarizing event lasted few seconds, the slow one several minutes. In some cells, only the slow depolarizing component was observed. During the slow depolarizing event, the sensitivity of the membrane potential for a change in the K + gradient decreased, indicating a decrease in the relative K + permeability of the membrane. The rapid SP-induced depolarization could be reversed, when the membrane potential was depolarized to about 0 mV by elevation of the extracellular K + concentration, indicating a reversal potential close to the Cl- equilibrium potential. When the membrane was damped close to the resting membrane potential using the whole-ceU patch-clamp technique, SP induced a biphasic inward current with a similar time course as the SP-induced membrane depolarization. Evaluating current-to-voltage curves indicated a conductance decrease during the slow inward current with a reversal potential of the SP-dependent current close to the K ÷ equilibrium potential. The mean open time of single K + channels, measured in the cell-attached configuration of the patch-clamp technique, decreased after application of SP. In contrast, the mean open time of single Cl- channels increased. We conclude that activation of neurokinin receptors in astrocytes modulates the activity of K + and CI- channels, leading to a complex depolarization of the membrane potential. INTRODUCTION In recent years, several n e u r o t r a n s m i t t e r receptors have been identified on glial cells 17. These receptors include those which are directly linked to ionic channels such as the G A B A and glutamate r e c e p t o r of astrocytes 6' 14,34 as well as those which are most likely linked to K ÷ channels via second messengers or G proteins such as the a - a d r e n e r g i c r e c e p t o r 11. In addition to the 'classical' n e u r o t r a n s m i t t e r receptors, glial cells also express receptors for n e u r o p e p t i d e s , e.g. receptors binding neurokinin agonists 5'35 and bradykinin 28. Recently, we could demonstrate that activation of neurokinin receptors leads to changes in the m e m b r a n e potential of cultured astrocytes from rat cortex 37. Brief application of neurokinin agonists, such as SP or neurokinin A , led to a rapid and transient d e p o l a r i z a t i o n followed by a long-lasting (several minutes) depolarization of the m e m b r a n e . The ionic mechanism underlying this depolarization was unidentified. In neurons and muscle cells, activation of SP receptors affects a variety of different m e m b r a n e channels. In bullfrog sympathetic neurons and toad smooth muscle cells, SP and acetylcholine inhibit M currents 1' 12,30. A c t i v a t i o n of SP receptors also exerts a direct effect
on acetylcholine-induced currents 8. In spinal dorsal horn cells of the rat, SP augments a calcium-sensitive current 21, while in guinea pig inferior mesenteric ganglion cells, b o t h decreases and increases in m e m b r a n e conductance were observed, which were not m e d i a t e d by M currents 9. A closure of K ÷ channels by activation of SP receptors was o b s e r v e d in spinal cord n e u r o n s 26 which is a c c o m p a n i e d by an increase in the N a ÷ p e r m e a b i l i t y in inferior mesenteric ganglion cells of the guinea pig 19. These results illustrate that n e u r o k i n i n r e c e p t o r s can be coupled to a variety of different ion channels, and that the set of channels influenced by neurokinins varies among different cell types. T h e link b e t w e e n SP r e c e p t o r activation and the conductance change was identified in rat pancreatic acinar ceils as an IP3-1inked calcium mobilization 31. In neurons from nucleus basalis a pertussis toxin-insensitive G p r o t e i n m e d i a t e d SP-induced inhibition of inward rectifying K ÷ channels 23. Cultured astrocytes express a variety o f m e m b r a n e channels including different types of K ÷ channels 4'2s'33, N a ÷ channels 3 and CI- channels 2'25. These channels could be involved in the SP-induced m e m b r a n e depolarization. The aim of the present study was t h e r e f o r e to identify the ionic mechanism underlying the SP-induced m e m b r a n e
* Present address: Pharmaceutical Research Department, E Hoffmann-La Roche, Grenzacherstrasse 124, CH-4002 Basel, Switzerland
Correspondence: H. Kettenmann, Department of Neurobiology, University of Heidelberg, Im Neuenheimer Feld 345, 6900 Heidelberg, F.R.G. 0006-8993/91/$03.50 © 1991 Elsevier Science Publishers B.V. (Biomedical Division)
104 depolarization in cultured astrocytes of the rat. We demonstrate that activation of neurokinin receptors led to a closure of K ÷ channels and the opening of C1channels.
MATERIALS AND METHODS
Cell culture Cultures of enriched astrocytes ( > 90%) were obtained from cerebral hemispheres of 0- to 2-day-old Sprague-Dawley rats, according to McCarthy and de Vellis TM with modifications described by Keilhauer et alJ 3, and maintained in culture for 3-6 weeks. We stained sister cultures to those used in the electrophysioiogical experiments with GFAP antibodies to label the astrocytes in that culture system. These cells could be easily distinguished from the contaminating macrophages/microglial cells based on morphological criteria.
Solutions Cultures were continuously superfused during the experiments. Within 1-3 s 90% of the bathing fluid could be exchanged. The standard salt solution for intracellular recordings contained (in mM): KCI 5.4, NaC1 116, NaH2PO4.H20 1.0, MgSO 4 0.8, CaCI 2 1.8, D-glucose 5.6, NaHCO 3 26.2, HEPES 10. The pH was adjusted to 7.4. In some experiments NaCi was equimolarly replaced by KC! to depolarize the cells by increasing extracellular K ÷ concentration ([K+]o) from 5.4 to 50 or 121 mM. To alter both the K ÷ and CIgradient, NaC1 was replaced by potassium acetate. For patch-clamp recordings, the standard salt solution contained (in mM): KCI 5.4, NaCI 116, NaH2POa.H20 1.0, MgCI 2 0.8, CaCI 2 1.8, D-glucose 11.1, NaHCO 3 26.2, HEPES 5. The pH was adjusted to 7.2. Substance P (10-5"M; Sigma, Munich, E R . G ) was added to the standard salt solutions and applied by bath perfusion.
lmracellular recording Cultures were maintained on the stage of an inverted microscope at about 25 °C. For recording the membrane potential, cells were penetrated by 2 M KCl-filled electrodes (resistance 40-80 MI2) with the aid of a step-motor driven micromanipulator 32. Data were amplified by a WPI 707 amplifier and transferred to a computer system.
Patch-clamp procedures Whole-cell and single-channel currents were recorded with an EPC-7 amplifier (List, Darmstadt, F.R.G.) using standard methods of the patch-clamp technique ~°. Pipettes for whole cell recordings contained (in raM): KCI 100, NaCI 30, MgCI 2 1, CaCI 2 0.5, EGTA 5, HEPES 10, ATP 3, GTP 0.1. For single-channel recording, pipettes contained: KCI 146, MgCI 2 1.2, CaC12 2.6, HEPES 10. The pH was adjusted to 7.2. To determine the reversal potential of the SP response, currents were measured in the whole-cell-clamp configuration during application of a voltage ramp. The voltage ramp ranged from -150 to 90 mV and lasted 1.6 s.
Data processing Data were digitized by an interface card connected to an AT-compatible computer. Data for single-channel analysis were filtered at 2 kHz (Bessel 8-pole) and sampled at a frequency of 5 kHz. A Pascal program package developed in our laboratory was used to display and analyze the recorded traces. The program contained routines to determine current amplitudes and open state probabilities of single-channel currents. In addition, the computer was connected to the voltage control of the patch-clamp amplifier and generated the voltage changes during recording.
RESULTS
SP-induced membrane depolarizations Application of SP for 10 s induced a depolarization of the membrane potential in cultured astrocytes from rat cortex (Fig. 1, inset). The mean depolarization was 4.5 mV (n = 6) and lasted for minutes as has been previously described 37. The depolarization was preceded in about half of the experiments by a rapid, transient depolarization of about 4 mV, lasting few seconds.To characterize the slow depolarization, we analyzed the K ÷ sensitivity of the membrane by altering the K ÷ gradient. When [K+]o was increased from 5.4 to 50 mM, the membrane
Vm [mV]
1~SP
-30 -40
. I0 mV
-50 -60 -7o
q 'l~sP 5
min
Fig. 1. Sensitivity of the membrane potential for a change in the K ÷ gradient after application of SP. During recording of membrane potential (I'm) from an astrocyte with a conventional microelectrode, extracellular K + concentration was repeatedly elevated from 5.4 mM to 50 mM for 45 s as denoted by bars. SP was applied for 10 s as indicated by arrow. The inset on top shows the time course of an SP-induced membrane depolarization without altering the K + gradient. Bars in the inset denote 10 mV and 1 min.
121 m M K +
Fig. 2. Reversal of the first transient depolarization. During recording of membrane potential (Vm) from an astrocyte with a conventional microelectrode, SP was applied for 10 s as indicated by arrow. Resting membrane potential was -70 mV. A magnification of the SP-induced response is shown in the inset. After 25 rain the membrane was depolarized by elevating [K+]o from 5.4 to 121 mM. When SP (arrow) was again applied under these depolarized conditions, a transient hyperpolarizing response was elicited. The inset on the right displays a magnified trace.
105
®
®
(~
[nil
contro.~
TITI II1=I11I-I it 1 II II I II/]'TI'I
I
SP
1 min
o
sP
I I
gt/go
I SP I
1.1 o
1
• ~ D
OD
E
ol
÷
o I
OOOiD
• Io
oo
D QOO
oo
0.8 0.7 Fig. 3. Changes in membrane conductance induced by SP. A: membrane currents from an astrocyte with a large and flattened morphological appearance were recorded using the patch-clamp technique in the whole-cell recording configuration. The membrane potential was clamped at --60 mV and was repetitively switched for 400 ms intervals to -90 mV. SP was applied as indicated by a dashed line. The relative conductance (conductance at the indicated time (g0/conductance at the beginning of the recording (go)) is displayed corresponding to the current on top. B: a similar recording as described in the legend to 3A was obtained from a smaller cell (soma diameter less than 10/~m). SP did not elicit a detectable response. The top trace displays the current recording, the bottom trace the relative input conductance.
potential of astrocytes depolarized by about 50 mV, as previously described 2°. The size of the depolarization reflects the relative K ÷ conductance. After application of SP, K÷-induced depolarizations decreased in size, indicating a decrease of the relative K ÷ conductance (Fig. 1; n = 5). We analyzed the fast depolarizing event with an experimental paradigm previously applied to characterize G A B A - a c t i v a t e d CI- channels in astrocytes 15. SP was, therefore, applied under normal conditions and after depolarizing the membrane close to 0 mV by increasing [K+]o to 121 m M (Fig. 2). In elevated [K+]o, SP induced a hyperpolarizing response (1 mV, S.D. 0.6, n = 7) with a time course similar to the fast depolarization seen in normal [K+]o . To shift the CI- gradient to a more positive value under these depolarized conditions, 116 mM NaCI was replaced by an equimolar amount of potassium acetate ([Cl-]o was at 10 raM). The membrane potential again depolarized close to 0 mV in high [K+]o, but no response could be observed after administration of SP (not shown; n = 4). These experiments suggest an increase of a CI- conductance during the rapid depolarizing event. SP-induced currents in cultured astrocytes
We analyzed whole cell currents with the patch-clamp technique to determine the changes in conductance induced by SP, when the membrane was clamped close to
0_J___ _ ~.............. l
-120
Vm -60
t~ [mV]
Fig. 4. Reversal potential of the slow SP response. A: currentvoltage (l-V) curves were obtained by applying a voltage ramp changing the membrane potential (Vrn) from -120 to 0 mV while recording current (/) in the whole-cell recording mode of the patch-clamp technique. The I-V curve prior to application of SP (control) and 3 min after application of SP at the peak of the slow inward current (SP) are shown. B: the I - V curve of the SP-sensitive current was obtained by subtracting the 1-V curve after application of SP from that of the control (SP-control). The potassium equilibrium potential (EK+) is indicated by the dashed vertical line.
the resting potential. In contrast to the previous study with conventional electrodes aT, these recordings resolved two populations of cells with respect to SP sensitivity: about half of the cells responded with a change in the resting current to application of SP (Fig. 3A), while another population was unresponsive (Fig. 3B). These unresponsive cells (n = 22) had small somata ( < 10/tm) and were further characterized by a high input resistance (R i = 848 Mg2; range 40-4300 MI2). These cells were too small to be successfully impaled with conventional electrodes. In contrast, responsive cells (n = 23) had a large and flattened morphological appearance, a much lower input resistance R i = 168 Mg2 (range 5-1090 MI2; n = 16) and could be impaled with conventional microelectrodes. In some of the responsive cells (n = 7), the SP-induced current had a monophasic time course with a mean peak of 250 p A (range 5-1000 pA). In all other responsive cells (n = 16) the inward current was biphasic; a fast transient inward current (peak at 438 p A ; range 100-1500 pA) that inactivated within 30-60 s was followed by a slow and longer lasting inward current. The latter current peaked after about 2-5 min with an average amplitude of 534 p A (range 30-2800 pA). It slowly returned, but commonly did not reach the previous level even after several minutes. Both types of cells were GFAP-positive when sister cultures were immunolabelled.
106
-30 I[pA]
t I, V _ I [ r n V ] ~ " " -60
,
i
5e 100 ms
SP+7 min
-~6
-
=1 IS . 10(1 m s
®
® PO
4 0 ~
[*/*l 20
0
10
20 rain
Fig. 5. Recordings from K÷ channels. A: a current (/) to voltage (V) curve was obtained from single channel recordings with the patch-clamp technique in the cell-attached configuration. The insets show two sample recordings at -18 and -78 mV (calibration bars refer to the inset traces). The K+ (EK+) and CI- (EcF) equilibrium potentials are indicated by the dashed lines. The extrapolated reversal potential of the current is close to EK+ identifying this channel as a K + channel. B: this scheme illustrates the electrode arrangement in the cell attached recording configuration. SP was applied to the part of the membrane which was not covered by the pipette. C: sample traces of current recordings from the K ÷ channels shown in A prior to SP application (control) and 7 (SP+7 min) and 23 min (SP+23 min) after SP application. The pipette contained 146 mM K ÷ and 154 mM CF. D: from intact patch recordings of 26 s duration, the mean open state probability of the first current level (Po) was determined before and at different times after a 10 s SP application (indicated by the dashed line).
Effect of SP on cell input conductance To determine the conductance changes during an SP response, we recorded currents at two different membrane potentials by briefly changing the holding potential (-60 mV) to a more negative value (-90 mV). These potential jumps induced inward currents. After applying SP, the amplitude of these currents decreased in most cells (Fig. 3A). This finding indicates that the slow component of the SP response is accompanied by a
I [pA]
(~
control
[1 pA 100 ms
SP + 18 rain
I V
-150 -100 -5{2
50
[mV]
-2
®
conductance decrease. It was not possible to resolve conductance changes during the fast, transient response and all subsequent data were obtained at the peak of the slow, sustained phase. To further characterize these conductance changes, we determined current-to-voltage curves of the membrane before and after application of SP. The membrane potential of astrocytes was continuously changed between -150 and 90 mV within 1.6 s by a voltage-ramp to determine steady-state current-voltage (I-V) relations. I - V curves were recorded before (control) and 2 min after application of SP (10-5 M) (at the peak of the slow, sustained response), and the slopes representing cell input conductance were compared. During the application of SP the input conductance decreased during the slow, sustained phase in a similar way as has been observed from recordings with rectangular current pulses (Fig. 4). From these two approaches we determined the mean input conductance changes. During the slow, sustained phase the input conductance showed a decrease to 95% (S.D. = 5%; n = 15) indicating the closure of ion channels during that phase. In many cases we observed that the increase in input resistance was not fully reversible and continued to last for the rest of the recording time (up to more than one hour). Under these conditions, a second application of SP induced only a very small or even no detectable current response.
Reversal potential of the SP-sensitive current The reversal potential of the SP-dependent current was determined to characterize the ionic species mediating the SP-induced changes. We could only analyze the slow component of the SP effect, since the inactivation of the fast, transient component was too fast to obtain the I - V relation. I - V curves at the peak of the slow inward current response were subtracted from those obtained under control conditions. This subtraction yielded the current component which was blocked after application of SP (Fig. 4). The mean reversal potential of this current was -93 mV (S.D. 25 mV, n = 5) and thus close to the K ÷ equilibrium potential (EK = -81 mV) indicating that K ÷ channels close during the slow, sustained phase.
Possible effect of SP on M-type K + channels
® [v.]
40
30 20 10 0
10
20 min
Fig. 6. Single channel recordings from a CI- channel. A. CI -selective channel was identified by its reversal potential close to the Cl--equilibrium potential. For details see legend to Fig. 5.
Several groups have reported that the action of neurokinins on neuronal membranes is mediated via a blockage of M-type K + channels 1'26. We applied the voltage-clamp protocol designed to isolate M currents 1 before and during application of SP. The membrane was clamped at depolarized membrane potentials (-15 or -35 mV) and hyperpolarizing voltage steps close to the resting potential (-60 mV) were applied for 0.4-2 s.
107
Fig. 7. Scheme of neurokinin receptor involvement in [K+]o homeostasis. K÷ is released during electrical activity of a neuron (N; series of action potentials displayed in the inset). Astrocytes (A) take up and redistribute the excess K+ from the extraceUular space. Local activation of glial neurokinin receptors alter the direction of the K÷ currents. For details see discussion.
Under these conditions, no relaxation of an outward current was observed during or after application of SP (not shown; n = 6). We therefore conclude that astrocytes under our culture conditions did not express M-type K ÷ channels which are affected by SP.
Identification of K ÷ and CI- channels Single channel recordings were used to further characterize the interaction between neurokinin receptor activation and its effect on electrical membrane properties. Before or after testing the effect of SP on the single-channel properties, the reversal potential was determined to identify the ionic species permeating the channel. To compare reversal potentials with the equilibrium potentials for K + and Cl-, we made the following assumptions: the resting membrane potential was assumed at -63 mV (mean value determined at the beginning of whole-cell recordings; S.D. 5 mV, n = 26), intracellular [K +] at 100 mM 36, intracellular [Cl-] at 40 mM 15. When using 146 mM K + and 154 mM Cl- in the pipette for cell-attached recordings, the K + equilibrium potential was at 10 mV, the Cl- equilibrium potential at -34 mV. Voltage ramps ranging from 57 mV to -183 mV (membrane potential values based on a mean resting potential as described above) were applied to determine the reversal potentials of single-channel currents (not shown). In other experiments, we determined the reversal potential from continuous recordings at different membrane potentials (see Figs. 5A and 6A). We identified two populations of channels, one population with a reversal potential close to the K + equilibrium potential (n = 29), and one with a reversal potential close to the Clequilibrium potential (n = 5).
Effect of SP on single-channel properties While recording from single channels in the cellattached configuration of the patch-clamp technique (Fig. 5B), neurokinin receptors were activated by applying SP (10-5 M) to the bath for 10 s. We analyzed mean open time and the decay time constant of open and closed time histograms of the single channel properties before, during and after application of SP. In 8 out of 10 experiments, the mean open time of K ÷ channels determined from recordings of 26 s duration, decreased to 12% (S.D. 9.5%) of the control value within the first 7 min (Fig. 5C,D). The peak of the decrease in mean open time occurred after less than 2 min in 5 patches and after 6-7 min in the other 3 recordings. In 2 recordings we observed the opposite, namely an increase of the mean open time to 250% and 2300% of the control value. We selected 3 recordings which showed a decrease in open probability after application of SP and analyzed the decay constants of the open and closed times. These patches did not show any superposition of simultaneous openings. In these samples, the decay constant of the open times decreased to 64% of the control value, while the two decay constants of the closed times increased to 108% and to 250% of the control, respectively. The amplitude of the single-channel current was slightly reduced after application of SP; this was most likely due to the SP-induced depolarization.We were able to test the effect of SP on the mean open time of 2 out of 5 identified CIchannels (Fig. 6C,D). In both cases, mean open probability increased 9- and 15-fold, respectively, within 3 min. This increase was in part reversible. DISCUSSION
Neurokinin receptor activation and ion channel link In this study we provide evidence that the slow membrane depolarization in astrocytes is caused by an increased tendency of K ÷ channels to close. This can be deduced from the finding that the relative K + permeability during the depolarized phase decreased in concert with a decrease in membrane conductance. This observed conductance decrease is small, but it should be taken into account that we recorded from a large syncytium and that a large proportion of the injected current did spread to other cells via gap junctions which are numerously present in these cells 16. Additional evidence comes from recordings of single K ÷ channels. We observed that K ÷ channels were decreased in their open state probability after neurokinin receptor activation. This decrease is not due to membrane depolarization per se, since the mean open time of most K + channels in astrocytes increased with membrane depolarization. A more detailed characterization of the different types of K + channels in
108 astrocytes, however, is required to link the activation of neurokinin receptors to defined populations of K ÷ channels. The first transient depolarization which was observed in many, but not all astrocytes could be reversed by depolarizing the membrane potential in high K ÷ similar as has been shown for the G A B A activated CI- channel in astrocytes 15. This observation together with the notion that CI- channel open probability was increased after application of SP indicates that SP activates a C1conductance with a different time course as the decrease in K ÷ channel conductance. Both conductance changes are superimposed after activation of SP receptors and the different time courses could explain the complicated time courses of membrane depolarization 37 or inward currents. When recording from channels in the intact patch configuration, bath application of SP did not directly affect the channel under the pipette. We can therefore exclude a direct effect of SP on K ÷ or C1- channels and it is most likely that the link between SP receptor activation and channel modulation is mediated by second messengers e.g. by IP3-1inked calcium mobilization as described in rat pancreatic acinar cells31.
Comparison of glial and neuronal SP responses In neurons, brief application of neurokinin agonists induced a depolarization of the membrane potential similar to that described in this study for cultured astrocytes ~9. Neuronal neurokinin receptor activation can exert multiple and complex effects on the activity of different ion channels 9'22. This includes an inactivation of M currents 1 and an activation of a cation-permeability 29. In this study we did not observe changes in currents recorded according to an M-channel identification protocol after application of SP. The effect of neurokinin receptor activation on the CIconductance in astrocytes has so far not been described in any other cell system. It can therefore be concluded that neurokinin receptors are linked to specific astrocytic channels rather than that astrocytes express a neuronal-like type of neurokinin receptorqon channel system.
Heterogeneity of astrocytes with respect to SP responsiveness We found two populations of cells in the purified astrocytic cultures: large cells with a high input conductance which responded to SP. This population correREFERENCES 1 Adams, P.R., Brown, D.A. and Jones, S.W., Substance P inhibits the M-current in bullfrog sympathetic neurons, Br. J. Pharmacol., 79 (1983) 330-333. 2 Barres, B.A., Chun, L.L.Y. and Corey, D.P., Ion channel expression by white matter glia: I. Type 2 astrocytes and oligodendrocytes, Glia, 1 (1988) 10-30. 3 Bevan, S., Chiu, S.Y., Gray, P.T.A. and Ritchie, J.M., The
sponds to the electrically extensively coupled astrocytes which form a large syncytium in this culture system 16. The extent of electrical coupling is reflected in the high input conductance of these cells, a characteristic feature of the astrocytic syncytium. The smaller cells which were unresponsive to SP had a morphological appearance of more immature astrocytes which is reflected by their lower input conductance. It is likely that astrocytes express higher densities of SP receptors during maturation. This is also substantiated by our observations that SP responses increased in size with prolonged maintenance of cells in culture. Wilkin et al.38 also reported that SP receptor density increased with time in culture.
Functional implications of ion channel control by neurokinin receptors One of the important functions of astrocytes is the maintenance of the K + homeostasis in the extraceUular space. It was demonstrated recently that passive currents which redistribute K ÷ and thereby regulate [K+]o, can be facilitated by an uneven distribution of K + channels along glial cells 24. This allows a spatial control and enforces a direction of these currents. The modulation of K ÷ and CI- channels by neuropeptides could add a temporal component to this control. It can be speculated that in areas where SP receptors are activated and K ÷ channels are closed, K ÷ release and uptake are suppressed. Peptide receptors on astrocytes could therefore guide spatial buffer currents which could result in temporal changes in glial K ÷ regulation (Fig. 7). It has been demonstrated for Schwann cells that K ÷ channel activity is linked to the proliferative state of that ceil. Moreover, blockade of K ÷ channels led to a blockade of proliferation in the Schwann cells 7 and in retinal glial ceUs27. Assuming a similar link between proliferation and K ÷ channel activity in astrocytes, local levels of SP in the brain could regulate the proliferative state of these cells.
Acknowledgements. This research was supported by Bundesministerium fiir Forschung und Technologie and Deutsche Forschungsgemeinschaft (Heisenberg-Stipendium to H.K.I SFB 317). The authors thank Drs. H. Marrero, M. Schachner, J. Trotter, G. Trube and M. Wienrich for discussion and helpful comments on the manuscript, and T. Berger, S. Hauck, and S. Riese for excellent technical assistance. presence of voltage-gated sodium, potassium and chloride channels in rat cultured astrocytes, Proc. R. Soc. London, 225 (1985) 299-313. 4 Bevan, S. and Raft, M., Voltage-dependent potassium currents in cultured astrocytes, Nature, 315 (1985) 229-232. 5 Beaujouan, J.C., Daguet de Montety, M.D., Torrens, Y., Dietl, M. and Glowinski, J., Selective presence of a high density of NK-1 receptors on astrocytes from the rat brain stem in primary culture, Reg. Peptides, 22 (1988) 33.
109 6 Bormann, J. and Kettenmann, H., Patch clamp study of GABA receptor CI- channels in cultured astrocytes, Proc. Natl. Acad. Sci. U.S.A., 85 (1988) 9336-9340. 7 Chiu, S.Y. and Wilson, G.E, The role of potassium channels in Schwann cell proliferation in Wallerian degeneration of explant rabbit sciatic nerves, J. Physiol. (London), 408 (1989) 199-222. 8 Clapham, D.E. and Neher, E., Substance P reduces acetylcholine-induced currents in isolated bovine chromaffin cells, J. Physiol. (London), 347 (1984) 255-277. 9 Griffith, W.H., Hills, J.M. and Brown, D.A., Substance P-mediated membrane currents in voltage-clamped guinea pig inferior mesenteric ganglion cells, Synapse, 2 (1988) 432-441. 10 Hamill, O.P., Marty, A., Neher, E., Sakmann, B. and Sigworth F.J., Improved patch-clamp techniques for high-resolution current recording from cells and cell-free membrane patches, Pflagers Arch., 391 (1981) 85-100. 11 Hirata, H., Slater, N.T. and Kimelberg, H.K., Alpha-adrenergic receptor-mediated depolarization of rat neocortical astrocytes in primary culture, Brain Research, 270 (1983) 358-362. 12 Jones, S.W., Muscarinic and peptidergic excitation of bull-frog sympathetic neurones, J. Physiol. (London), 366 (1985) 63-87. 13 Keilhauer, G., Meier, D.H., Kuhlmann-Krieg, S., Nieke, J. and Schachner, M., Astrocytes support incomplete differentiation of an oligodendrocyte precursor cell, EMBO J., 44 (1985) 24992504. 14 Kettenmann, H., Backus, K.H. and Schachner, M., Aspartate, glutamate and gamma-aminobutyric acid depolarize cultured astrocytes, Neurosci. Lett., 52 (1984) 25-29. 15 Kettenmann, H., Backus, K.H. and Schachner, M., GABA opens Cl-channels in cultured astrocytes, Brain Research, 404 (1987) 1-9. 16 Kettenmann, H. and Ransom, B.R., Electrical coupling between astrocytes and between oligodendrocytes in mammalian cell cultures, Glia, 1 (1988) 64-73. 17 Kimelberg, H.K. (Ed.), Glial Cell Receptors, Raven Press, New York, 1988. 18 McCarthy, K.D. and De Vellis, J., Preparation of separate astroglial and oligodendroglial cell cultures from rat cerebral tissue, J. Cell Biol., 85 (1980) 890-902. 19 Minota, S., Dun, N.J. and Karczmar, A.G., Substance Pinduced depolarization in sympathetic neurons: not simple K-inactivation, Brain Research, 216 (1981) 224-228. 20 Moonen, G. and Nelson, P.G., Some properties of astrocytes in primary cultures. In E. Schoffeniels, G. Frank, L. Hertz and D.B.L. Tower (Eds.), Dynamic Properties of Glial Cells, Pergamon Press, London, 1978, pp. 389-393. 21 Murase, K., Ryu, P.D. and Randic, M., Substance P augments a persistent slow inward calcium-sensitive current in voltageclamped spinal dorsal horn neurons of the rat, Brain Research, 365 (1986) 369-376. 22 Murase, K., Ryu, P.D. and Randic, M., Tachykinins modulate multiple conductances in voltage-clamped rat spinal dorsal horn neurons, J. Neurophysiol., 61 (1989) 854-865. 23 Nakajima, Y., Nakajima, S. and Inoue, M., Pertussis toxin-
24 25 26 27 28
29 30
31
32
33 34 35
36
37 38
insensitive G protein mediates substance P-induced inhibition of potassium channels in brain neurons, Proc. Natl. Acad. Sci. U.S.A., 85 (1988) 3643-3647. Newman, E.A., Distribution of potassium conductance in mammalian Miiller (glial) cells: a comparative study, J. Neurosci., 7 (1987) 2423-2432. Nowak, L., Ascher, P. and Berwald-Netter, Y., Ionic channels in mouse astrocytes in culture, J. Neurosci., 7 (1987) 101-109. Nowak, L.M. and Macdonald, R.L., Substance P decreases a potassium conductance of spinal cord neurons in cell culture, Brain Research, 214 (1981) 416--423. Puro, D.G., Roberge, F. and Chan, C.-C., Retinal glial cells and ion channels: a possible link, lnvestigat. Ophthalmol. Visual Sci., 30 (1989) 521-529. Reiser, G. and Hamprecht, B., Bradykinin induces hyperpolarizations in rat glioma cells and in neuroblastoma x glioma hybrid cells, Brain Research, 239 (1982) 191-199. Reiser, G. and Hamprecht, B, Substance P and serotonin act synergistically to activate a cation permeability in a neuronal cell line, Brain Research, 479 (1989) 40-48. Sims, S.M., Walsh, J.V. and Singer, J.J., Substance P and acetylcholine both suppress the same K + current in dissociated smooth muscle cells, J. Physiol. (London), 251 (1986) C580-C587. Song, S.-Y., Iwashita, S., Noguchi, K. and Konishi, S., Inositol triphosphate-linked calcium mobilization couples substance P receptors to conductance increase in rat pancreatic acinar cell line, Neurosci. Lett., 95 (1988) 143-148. Sonnhof, U., Frrderer, R., Schneider, W. and Kettenmann, H., Cell puncturing with a step motor driven manipulator with simultaneous control of displacement, Pflagers Arch., 329 (1982) 295-300. Sonnhof, U. and Schachner, M., Single voltage-dependent K÷-channels in cultured astrocytes, Neurosci. Lett., 64 (1986) 241-246. Sontheimer, H., Kettenmann, H., Backus, K.H. and Schachner, M., Glutamate opens Na+/K + channels in cultured astrocytes, Glia, 1 (1988) 328-336. Torrens, Y., Beaujouan, J.C., Saffroy, M., Daguet de Montety, M.C., Bergstr6m, L. and Glowinski, J., Substance P receptors in primary cultures of cortical astrocytes from the mouse, Proc. Natl. Acad. Sci. U.S.A., 83 (1986) 9216-9220. Walz, W., Wuttke, W. and Hertz, L., Astrocytes in primary cultures: membrane potential characteristics reveal exclusive potassium conductance and potassium accumulator properties, Brain Research, 292 (1984) 367-374. Wienrich, M. and Kettenmann, H., Activation of substance P receptors leads to membrane potential responses in cultured astrocytes, Glia, 2 (1989) 155-160. Wilkin, G.P., Cholewinski, A.J., Reid, J.C. and Marriott, D.R., Peptide receptors on astrocytes: characterization and regional heterogeneity, ln: Differentiation and Functions of Glial Cells, Satellite Symposium of the 12th Biennal Meeting of the International Society for Neurochemistry, Rome, 1989, p. 36.
