- Email: [email protected]
Rapid extracellular pH transients related to synaptic transmission in rat hippocampal slices
Brain Research, 436 (1987) 352-356 Elsevier
352 BRE 13148
Rapid extracellular pH transients related to synaptic transmission in rat hippocampal slices O.A. Krishtal, Yu. V. Osipchuk, T.N. Shelest and S.V. Smirnoff Bogomoletz Institute of Physiology, Ukrainian Academy of Sciences, Kiev (U.S.S.R.)
(Accepted 9 June 1987) Key words: Brain slice; pH measurement; Synaptic transmission
Extracellular pH changes were measured in the rat hippocampal slices using thc pH-sensitive dye Phenol red. pH changes accompanied artificially evoked synaptic transmission in the dendrite area of dentate gyrus neurones and pyramidal neurones (CAj). Single electrical stimulation of presynaptic pathways produced a rapid acidic pH shift which was followed by a long-lasting alkaline one. The duration (nearly 10 m-) and amplitude of the acidic shift were closely related to the orthodromically evoked population excitatory postsynaptic potential. Population action potential, when elicited antidromically or in conditions of blocked synaptic transmission, did not produce any pH changes which are supposed to be specifically linked to the synaptic transmission.
INTRODUCTION Neuronal activity is accompanied by considerable changes in the composition of the cellular microenvironment. In addition to the changes in Na, K and Ca concentrations, variations of extracellular pH have been demonstrated in cerebellar cortex 5 and in hippocam~_,al slices of rat 12. It was shown in cerebellar cortex that repetitive electrical stimulation is accompanied by an initial alcaline shift of about 0.05 pH unit (nearly I s in duration), which is followed by a long-lasting acid shift of about (!.2 pH unit (nearly 30 s in duration), pH was measured in these experiments by using pH-sensitive microelectrodes whose slow response limited the time resolution of the data to about 1 s. In the experiments reported here we used an optical method to monitor extracellular pH changes in the brain slices with increased speed and sensitivity of detection. Using this method we found rapid changes in pH that correlated with excitatory postsynaptic potentials (EPSP). Since synaptic vesicles are acidified by proton
ATPases 8, we suggest that the acid extraceiiular transients are due to the secretion of stored protons and/or to the insertion of proton ATPases into the membrane bounding the synaptic cleft. METHODS AND MATERIALS The experiments were performed on transversely sectioned rat hippocampal slices. The 200-400 g m thick slices were cut by hand using safety razor blade as described elsewhere 4. The chamber for experiments was mounted on the stage of inverted micros-cope (Fig. 1). For optical measurements the lower surface of the submerged slice was kept adjacent to the transparent chamber's bottom. Normal artificial cerebrospinal fluid (ACSF), equilibrated with the O2-CO 2 ( 9 5 % - 5 % ) gas mixture, was constantly superfusing the chamber at a rate of 1 ml/min. It had the following composition (in mmol/l) at pH 7.4: NaCl 134, KCI 5, CaCl 2 1, MgCl 2 2, KH2PO 4 1, NaHCO 3 16 and glucose 20. The experiments were carried out at room temperature between 21 and 25 °C. It was demonstrated previously that at these
Correspondence: O.A. Krishtal, Bogomoletz Institute of Physiology, Ukrainian Academy of Sciences, Bogomoletz Str. 4, Kiev 252024, U.S.S.R.
0006-8993/87/$03.50© 1987Elsevier Science Publishers B.V. (Biomedical Division)
353
->
~
,\\L/
f
I
}A° J I
Pc
Fig. 1. Simplified diagram of the experimental set-up. A, hippocampal slice; B, stimulating electrodes; C, recording electrodes; D, inlet and E, outlet of ACSF; F, glass coverslip; G, plastic light guide; H, light source with condenser: I, oil immersion microscopic objective: J, glass absorbance filter; K, photodiode; L, current/voltage converter; M, amplifier; MX, analog multiplexor: ADC, analog/digital converter; PC, computer: DACO, digital/analog converter for pathway stimulation: DAC1, digital/analog converter for computerized zero adjust.
temperatures the slices are adequately oxygenated via the one surface9. The stimulating electrodes were made from insulated sharpened Ni/Cr wire with a tip diameter of 30/,m. Opto-electronicaily coupled stimulating circuit was used. The duration of stimulus was normally 0.5 ms and the current amplitude was 50 /~A. Extracellular recording of field potential (population EPSP and spikes) was performed by 1/~m tip diameter glass micropipettes filled with ACSF and connected to a high impedance amplifier with a Bessel fourth order low-pass filter (cut-off frequency 1 kHz). Extracellular pH changes were measured by using the pH-sensitive dye Phenol red. Even at a high concentration (2 x 10-4 mmoFl), this dye did not affect the synaptic transmission in the slice when tested either in the darkness or under illumination. The deprotonated form of Phenol red has an optical absorption maximum at 560 nm, and the protonated form has a much weaker absorption at 440 nm. Control measurements confirmed that the optical properties of Phenol red did not change when the concentration of calcium, magnesium and potassium ions in tt.,: ACSF varied in the range of 0-10 mmoFl. Using the cell culture (neuroblastoma C 1300), we checked whether Phenol red permeates across the cell plasma membrane and found that it did not. After performing all these tests we could expect that the selected indicator allowed one to measure actual extraceUular
pH changes. The optical absorption of the slice perfused by the ACSF with Phenol red at the concentration of 2 x 10-4 mM was measured by passing light through the flexible plastic light guide, immediately adjacent to the top surface of the slice. The tip diameter of the guide was 50/~m. The light source, 24 v!150 W tungsten halogen lamp with condenser and heat filter, was powered by a stabilized power unit (ripples less than 0.01%). An oi!-immersion microscopic objective (100x, numerical aperture 1.4) collected the light passing through the slice and a 560nm glass absorption filter and sent it to a low-noise silicone photodiode. The latter was coupled to a lownoise current-voltage converter and amplifier with computerized digital zero-adjusting circuit and a 4pole Bessel low-pass filter (cut-off frequency 1 kHz). The light and field potential signals were digitized by a 12-bit analog-digital converter and averaged over 0-100 trials to improve the signal-to-noise ratio under control of an IBM PC-compatible computer. Optically detected pH changes were calibrated by measuring the changes in the light absorption when the slices were successively perfused by ACSF with different buffered pH levels. The equilibration of optical signal was achieved within 5 min, required for the diffusion of acidic or basic solutions to the whole volume of the slice. Usually pH change of 5 x 10-3 unit corresponded to a 0.01% change in absorption. Simultaneous recording of absorption at the wavelength corresponding to the isobestic point of the dye can be used to exclude possible artifacts due to the movement of preparation or changes in its volume li. Control measurements with 510 nm absorption filter (near the Phenol red isobestic point) revealed neglible signals. So we did not perform simultaneous recordings at the dye isobestic point in routine experiments. RESULTS
Two series of experiments were performed. In the first series the hippocampal perforant path was stimulated, while the recording electrode was placed near the somata of the dental gyrus granule cells, pH-sensitive dye absorption was measured in the area of granule cells somata and dendrites. In the other series the CA! pyramidal cells were stimulated via the Schaffer/commissural afferents in the stratum radia-
354
5qo'-Nl a)
"
_l
/
~
,--.~_ 15.10"3pH
1_
1_-" :': ..
~..-.,~.
/"
b)
2mYL 20ms
.. • "...,...
Ialkaline ~acid ----------__...__.
_
. • "~,.........%.
.
j,"'--- ~---.~
a second slow wave of acidification. It could be elicited by a burst of stimulating pulses (i.e. 10 pulses at 10 Hz) and lasted for more then 10 s. Both slow waves apparently correspond to the previously described pH transients 5, while the initial acidous transient is a new finding. All detected pH-transient waves were abolished when the slices were perfused with the O2-deprived medium, suggesting that the changes seen are not artifactual. Possible responsibility of N a + - H + and/or C1--HCO~ exchange for the detected p H changes have been examined by adding amiloride ( 5 0 / a M ) and DIDS (500 p M ) to the ACSF solution. Amiloride, an N a + - H + exchange blocker, had no effect on
15.,O%H
fL f"-.
.,_,-AlFig. 2. a: simultaneous recording of orthodromically evoked population EPSP and population spike (top) and pH changes shown in two different time scales (bottom) in hippocampal pyramidal CA l neurones; 50 optical responses were averaged. The stimulation frequency was 0.5 Hz. Alkaline shift corresponds to the upward deflection here and in further figures, b: field potential (~op) and pH recording (bottom) in CA~ region elicited by a pail" of orthodromic stimulae. The control optical response elicited by a single stimulus is shown by a dotted line. Arrows indicate the increased amplitude of relative acidification corresponding to the second stimulus. Note the different time scales in electrical and optical recordings.
2mV l
: ~ i
20ms 5.10.3 pH
f
---__
[
--'-":
~"
100 ms
2mV [
i
5.10.3pH [
20 ms lOOms
x
2 mV ] 20ms
rum, while extracellular field potential was recorded from the area of pyramidal cell somata and the pH was measured in the area of somata and dendrites of these cells. Electrical stimulation of the indicated pathways produced a synchronous monosynaptic activation of granule or pyramidal cells: population EPSP and population spikes were detected. We have found that orthodromically evoked synaptic transmission was accompanied by a series of pH changes (Fig. 2). A transient acid shift occurred immediately after electrical stimulus. It was followed by a much longer alkaline shift. The acid transient and population EPSP lasted for about le--20 ms, whereas the alkaline shift was much longer and lasted for about 1 s. The recovery of pH changes due to a single stimulus was completed within 2 s. Slower sweeps allowed us to detect
lOOms
2mV [ ,,.-"---"'-
d) J.
-
20 m s 5"IO'3pH [ 100 ms
Fig. 3. Effect of Mg2+ on the synaptic transmission and the extracellular pH shifts in the area of somata and dendrites of CA l pyramidal cells. The responses were induced by orthodromic stimulation of the Schaffer/commissural afferents in the stratum radiatum. Field potential (top) and pH recording (bottom) before (a) and after (b) the addition of 10 mmol/l Mg2+ to ACSF. c: recovery in normal ACSF. The stimulation frequency was 0.5 Hz; 50 optical responses were averaged, d: field potential and optical monitored pH changes in the dendrite area of the dentate gyrus granule cells antidromically stimulated via their axons; 200 cycles of averaging.
355 the detected pH changes. DIDS, the blocker of several anion exchangers 1°, when applied in a very high concentration of 500/~M, enhanced the alkalinization phase, but did not affect the initial acidification. The latter process was also unaltered under a carboanhydrase inhibitor acetazolamide (100~uM). It is well-known that Mg 2+ blocks synaptic transmission by inhibiting presynaptic transmitter release. It is effective in the concentrations that do not inhibit the propagation of action potentials in the presynaptic axons 3. After superfusion of the slices with ACSF containing 10 mmol/l Mg 2+ for 7 min, both electrically evoked field potentials and optically measured pH changes completely disappeared (Fig. 3). This effect was reversible and indicated that the action potentials in presynaptic fibers were not responsible for the detected pH changes. Antidromic stimulation of the postsynaptic dental gyrus cells and pyramidal CA~ cells did not produce measurable pH changes even at an increased number of averagings. So the action potentials in the somata of postsynaptic cells are not responsible for the pH changes as well. Consequently, we suppose that the pH changes are directly linked with the synaptic transmission. Using the effects of synaptic potentiation 13 to modulate the efficiency of the synaptic transmission in the slice, we have found that the value of the initial acidic shift is closely related to the amplitude of the population spike (Fig. 2), thus confirming the above assumption. DISCUSSION It was shown that the contents of the synaptic vesicles from the rat brain have a low pH s. This is caused by a proton transiocating ATPase in the membrane 12. The release of the acid contents of the vesicles during synaptic transmission may lead to the rapid acidic
REFERENCES 1 Cannon, C., van Adeisberg, J., Kelly, S. and AI-Awqati, Q., Carbon-dioxide-induced exocytotic insertion of H ÷ pumps in turtle-bladder luminal membrane: role of cell pH and calcium, Nature (London), 313 (1985) 443-446. 2 Gluck, S., Cannon. C. and AI-Awqati, Q., Excocytosisregulates H* transport in the turtle bladder by rapid insertion of H ÷ pumps into the luminal membrane, Proc. Natl. Acad. Sci. U.S.A., 79 (1982) 4327-4331. 3 Hacket, J.T., Selective antagonism of frog cerebeilar synaptic transmission by manganese and cobalt ions, Brain Re-
transient in the synaptic gap and its vicinity. The other possibility is that synaptic vesicles insert their proton ATPases into the nerve terminal on fusion. These proton pumps can now 'turn over' and reduce the synaptic cleft pH. Similar studies in epithelia have shown that exocytosis of acid vesicles causes insertion of proton pumps into the membrane with a consequent increase in proton transport across the membrane 1'2. Exocytotic fusion of synaptic vesicles is restricted to defined domains of the terminal. In such a case the described changes in pH should occur in the immediate vicinity of the synaptic junction and their local amplitudes should be much greater than the detected values. The cause of the delayed alkalinization requires further study. It has been shown that a considerable part of isolated neurones from the rat brain possess excitatory desensitizing responses activated by rapid drops in external pH ~4. Properties of the proton-activated Na conductance mechanism in these neurones are similar to those previously described for the rat sensory, neurones 6. The functional significance of such 'proton receptor" is so far unclear (see also ref. 7). But since this mechanism is sensitive to rapid extraceilular pH changes, it can produce depolarization of the neuronal membrane in response to the transient pH drops described in this investigation and thus contribute to the interneuronai communications. Irrespective of the validity of this assumption, the complex sequence of rapid and slow pH changes accompanying synaptic transmission is a new factor which may be relevant and important for the function of the brain. ACKNOWLEDGEMENTS The authors are grateful to Prof. Q. AI-Awqati for valuable comments on the manuscript.
search, 114 (1976) 47-52. 4 Kerkut, G.A. and Wheal, H.V., (Eds.), Electrophysiology of Isolated blammalian CNS Preparation, Academic, London, 1981. 5 Kraig, P.R., Ferreira-Fiiho, C.R. and Nicholson, C., Alkaline and acid transients in cerebellar microenvironment, J. Neurophysiol., 49 (1983) 831-850. 6 Krishtal, O.A. and Pidoplichko, V.I., A receptor for protons in the nerve cell membrane, Neuroscience, 5 (1980) 2325-2327. 7 Krishta!, O.A. and Pidioplichko. V.I., A receptor for proton in the membrane of sensory neurones may participate in
356 nociception, Neuroscience, 6 ( 1981) 2599-2601. 8 Melnik, V.I. and Glebow, R.N., Kryhanovski, G.N., ATPdependent translocation of protons across the membrane of rat brain synaptic vesicles, Bull. Exp. Biol. Med., 99 (1985) 35-38. 9 Nicholson, J. and Honnsquard, N., Diffusion in the slice microenvironment for physiological studies, Fed. Proc., 42 (1983) 2865-2868. 10 Passow, H., Fasold, H., Jennings, M.L. and Lepke, S., The study of anion transport protein ('Band 3 Protein') in the rat cell membrane by means of tritiated DIDS. In J.A. Zadunaisky (Ed.), Cloride Transport in Biological Membranes, Academic, New York, I982, pp. 1-31.
11 Piwnica-Worms, D. and Lieberman, M., Microflourometric monitoring of pH l in cultured heart cells: Na+-H ÷ exchange, Am. J. Physiol., 244 (Cell Physiol., 13) (1983) c422-428. 12 Stadler, H. and Tsukita, S., Synaptic vesicles contain an ATP-dependent proton pump and show 'knob-like' protrusions on their surface, EMBO J., 3 (1984) 3333-3337. 13 Voronin, I., Long-term potentiation in the hippocampus slices, Neuroscience, 10 (1983) 1051-1069. 14 Vrublevski, S.V., Krishtal, O.A., Osipchuk, Yu.V. and Tsyndrenko, A.Ya., Proton-activated Na-conduction in cerebral cortex neurones, Dokl. Akad. Nauk SSSR, 284 (1985) 990-992.
352 BRE 13148
Rapid extracellular pH transients related to synaptic transmission in rat hippocampal slices O.A. Krishtal, Yu. V. Osipchuk, T.N. Shelest and S.V. Smirnoff Bogomoletz Institute of Physiology, Ukrainian Academy of Sciences, Kiev (U.S.S.R.)
(Accepted 9 June 1987) Key words: Brain slice; pH measurement; Synaptic transmission
Extracellular pH changes were measured in the rat hippocampal slices using thc pH-sensitive dye Phenol red. pH changes accompanied artificially evoked synaptic transmission in the dendrite area of dentate gyrus neurones and pyramidal neurones (CAj). Single electrical stimulation of presynaptic pathways produced a rapid acidic pH shift which was followed by a long-lasting alkaline one. The duration (nearly 10 m-) and amplitude of the acidic shift were closely related to the orthodromically evoked population excitatory postsynaptic potential. Population action potential, when elicited antidromically or in conditions of blocked synaptic transmission, did not produce any pH changes which are supposed to be specifically linked to the synaptic transmission.
INTRODUCTION Neuronal activity is accompanied by considerable changes in the composition of the cellular microenvironment. In addition to the changes in Na, K and Ca concentrations, variations of extracellular pH have been demonstrated in cerebellar cortex 5 and in hippocam~_,al slices of rat 12. It was shown in cerebellar cortex that repetitive electrical stimulation is accompanied by an initial alcaline shift of about 0.05 pH unit (nearly I s in duration), which is followed by a long-lasting acid shift of about (!.2 pH unit (nearly 30 s in duration), pH was measured in these experiments by using pH-sensitive microelectrodes whose slow response limited the time resolution of the data to about 1 s. In the experiments reported here we used an optical method to monitor extracellular pH changes in the brain slices with increased speed and sensitivity of detection. Using this method we found rapid changes in pH that correlated with excitatory postsynaptic potentials (EPSP). Since synaptic vesicles are acidified by proton
ATPases 8, we suggest that the acid extraceiiular transients are due to the secretion of stored protons and/or to the insertion of proton ATPases into the membrane bounding the synaptic cleft. METHODS AND MATERIALS The experiments were performed on transversely sectioned rat hippocampal slices. The 200-400 g m thick slices were cut by hand using safety razor blade as described elsewhere 4. The chamber for experiments was mounted on the stage of inverted micros-cope (Fig. 1). For optical measurements the lower surface of the submerged slice was kept adjacent to the transparent chamber's bottom. Normal artificial cerebrospinal fluid (ACSF), equilibrated with the O2-CO 2 ( 9 5 % - 5 % ) gas mixture, was constantly superfusing the chamber at a rate of 1 ml/min. It had the following composition (in mmol/l) at pH 7.4: NaCl 134, KCI 5, CaCl 2 1, MgCl 2 2, KH2PO 4 1, NaHCO 3 16 and glucose 20. The experiments were carried out at room temperature between 21 and 25 °C. It was demonstrated previously that at these
Correspondence: O.A. Krishtal, Bogomoletz Institute of Physiology, Ukrainian Academy of Sciences, Bogomoletz Str. 4, Kiev 252024, U.S.S.R.
0006-8993/87/$03.50© 1987Elsevier Science Publishers B.V. (Biomedical Division)
353
->
~
,\\L/
f
I
}A° J I
Pc
Fig. 1. Simplified diagram of the experimental set-up. A, hippocampal slice; B, stimulating electrodes; C, recording electrodes; D, inlet and E, outlet of ACSF; F, glass coverslip; G, plastic light guide; H, light source with condenser: I, oil immersion microscopic objective: J, glass absorbance filter; K, photodiode; L, current/voltage converter; M, amplifier; MX, analog multiplexor: ADC, analog/digital converter; PC, computer: DACO, digital/analog converter for pathway stimulation: DAC1, digital/analog converter for computerized zero adjust.
temperatures the slices are adequately oxygenated via the one surface9. The stimulating electrodes were made from insulated sharpened Ni/Cr wire with a tip diameter of 30/,m. Opto-electronicaily coupled stimulating circuit was used. The duration of stimulus was normally 0.5 ms and the current amplitude was 50 /~A. Extracellular recording of field potential (population EPSP and spikes) was performed by 1/~m tip diameter glass micropipettes filled with ACSF and connected to a high impedance amplifier with a Bessel fourth order low-pass filter (cut-off frequency 1 kHz). Extracellular pH changes were measured by using the pH-sensitive dye Phenol red. Even at a high concentration (2 x 10-4 mmoFl), this dye did not affect the synaptic transmission in the slice when tested either in the darkness or under illumination. The deprotonated form of Phenol red has an optical absorption maximum at 560 nm, and the protonated form has a much weaker absorption at 440 nm. Control measurements confirmed that the optical properties of Phenol red did not change when the concentration of calcium, magnesium and potassium ions in tt.,: ACSF varied in the range of 0-10 mmoFl. Using the cell culture (neuroblastoma C 1300), we checked whether Phenol red permeates across the cell plasma membrane and found that it did not. After performing all these tests we could expect that the selected indicator allowed one to measure actual extraceUular
pH changes. The optical absorption of the slice perfused by the ACSF with Phenol red at the concentration of 2 x 10-4 mM was measured by passing light through the flexible plastic light guide, immediately adjacent to the top surface of the slice. The tip diameter of the guide was 50/~m. The light source, 24 v!150 W tungsten halogen lamp with condenser and heat filter, was powered by a stabilized power unit (ripples less than 0.01%). An oi!-immersion microscopic objective (100x, numerical aperture 1.4) collected the light passing through the slice and a 560nm glass absorption filter and sent it to a low-noise silicone photodiode. The latter was coupled to a lownoise current-voltage converter and amplifier with computerized digital zero-adjusting circuit and a 4pole Bessel low-pass filter (cut-off frequency 1 kHz). The light and field potential signals were digitized by a 12-bit analog-digital converter and averaged over 0-100 trials to improve the signal-to-noise ratio under control of an IBM PC-compatible computer. Optically detected pH changes were calibrated by measuring the changes in the light absorption when the slices were successively perfused by ACSF with different buffered pH levels. The equilibration of optical signal was achieved within 5 min, required for the diffusion of acidic or basic solutions to the whole volume of the slice. Usually pH change of 5 x 10-3 unit corresponded to a 0.01% change in absorption. Simultaneous recording of absorption at the wavelength corresponding to the isobestic point of the dye can be used to exclude possible artifacts due to the movement of preparation or changes in its volume li. Control measurements with 510 nm absorption filter (near the Phenol red isobestic point) revealed neglible signals. So we did not perform simultaneous recordings at the dye isobestic point in routine experiments. RESULTS
Two series of experiments were performed. In the first series the hippocampal perforant path was stimulated, while the recording electrode was placed near the somata of the dental gyrus granule cells, pH-sensitive dye absorption was measured in the area of granule cells somata and dendrites. In the other series the CA! pyramidal cells were stimulated via the Schaffer/commissural afferents in the stratum radia-
354
5qo'-Nl a)
"
_l
/
~
,--.~_ 15.10"3pH
1_
1_-" :': ..
~..-.,~.
/"
b)
2mYL 20ms
.. • "...,...
Ialkaline ~acid ----------__...__.
_
. • "~,.........%.
.
j,"'--- ~---.~
a second slow wave of acidification. It could be elicited by a burst of stimulating pulses (i.e. 10 pulses at 10 Hz) and lasted for more then 10 s. Both slow waves apparently correspond to the previously described pH transients 5, while the initial acidous transient is a new finding. All detected pH-transient waves were abolished when the slices were perfused with the O2-deprived medium, suggesting that the changes seen are not artifactual. Possible responsibility of N a + - H + and/or C1--HCO~ exchange for the detected p H changes have been examined by adding amiloride ( 5 0 / a M ) and DIDS (500 p M ) to the ACSF solution. Amiloride, an N a + - H + exchange blocker, had no effect on
15.,O%H
fL f"-.
.,_,-AlFig. 2. a: simultaneous recording of orthodromically evoked population EPSP and population spike (top) and pH changes shown in two different time scales (bottom) in hippocampal pyramidal CA l neurones; 50 optical responses were averaged. The stimulation frequency was 0.5 Hz. Alkaline shift corresponds to the upward deflection here and in further figures, b: field potential (~op) and pH recording (bottom) in CA~ region elicited by a pail" of orthodromic stimulae. The control optical response elicited by a single stimulus is shown by a dotted line. Arrows indicate the increased amplitude of relative acidification corresponding to the second stimulus. Note the different time scales in electrical and optical recordings.
2mV l
: ~ i
20ms 5.10.3 pH
f
---__
[
--'-":
~"
100 ms
2mV [
i
5.10.3pH [
20 ms lOOms
x
2 mV ] 20ms
rum, while extracellular field potential was recorded from the area of pyramidal cell somata and the pH was measured in the area of somata and dendrites of these cells. Electrical stimulation of the indicated pathways produced a synchronous monosynaptic activation of granule or pyramidal cells: population EPSP and population spikes were detected. We have found that orthodromically evoked synaptic transmission was accompanied by a series of pH changes (Fig. 2). A transient acid shift occurred immediately after electrical stimulus. It was followed by a much longer alkaline shift. The acid transient and population EPSP lasted for about le--20 ms, whereas the alkaline shift was much longer and lasted for about 1 s. The recovery of pH changes due to a single stimulus was completed within 2 s. Slower sweeps allowed us to detect
lOOms
2mV [ ,,.-"---"'-
d) J.
-
20 m s 5"IO'3pH [ 100 ms
Fig. 3. Effect of Mg2+ on the synaptic transmission and the extracellular pH shifts in the area of somata and dendrites of CA l pyramidal cells. The responses were induced by orthodromic stimulation of the Schaffer/commissural afferents in the stratum radiatum. Field potential (top) and pH recording (bottom) before (a) and after (b) the addition of 10 mmol/l Mg2+ to ACSF. c: recovery in normal ACSF. The stimulation frequency was 0.5 Hz; 50 optical responses were averaged, d: field potential and optical monitored pH changes in the dendrite area of the dentate gyrus granule cells antidromically stimulated via their axons; 200 cycles of averaging.
355 the detected pH changes. DIDS, the blocker of several anion exchangers 1°, when applied in a very high concentration of 500/~M, enhanced the alkalinization phase, but did not affect the initial acidification. The latter process was also unaltered under a carboanhydrase inhibitor acetazolamide (100~uM). It is well-known that Mg 2+ blocks synaptic transmission by inhibiting presynaptic transmitter release. It is effective in the concentrations that do not inhibit the propagation of action potentials in the presynaptic axons 3. After superfusion of the slices with ACSF containing 10 mmol/l Mg 2+ for 7 min, both electrically evoked field potentials and optically measured pH changes completely disappeared (Fig. 3). This effect was reversible and indicated that the action potentials in presynaptic fibers were not responsible for the detected pH changes. Antidromic stimulation of the postsynaptic dental gyrus cells and pyramidal CA~ cells did not produce measurable pH changes even at an increased number of averagings. So the action potentials in the somata of postsynaptic cells are not responsible for the pH changes as well. Consequently, we suppose that the pH changes are directly linked with the synaptic transmission. Using the effects of synaptic potentiation 13 to modulate the efficiency of the synaptic transmission in the slice, we have found that the value of the initial acidic shift is closely related to the amplitude of the population spike (Fig. 2), thus confirming the above assumption. DISCUSSION It was shown that the contents of the synaptic vesicles from the rat brain have a low pH s. This is caused by a proton transiocating ATPase in the membrane 12. The release of the acid contents of the vesicles during synaptic transmission may lead to the rapid acidic
REFERENCES 1 Cannon, C., van Adeisberg, J., Kelly, S. and AI-Awqati, Q., Carbon-dioxide-induced exocytotic insertion of H ÷ pumps in turtle-bladder luminal membrane: role of cell pH and calcium, Nature (London), 313 (1985) 443-446. 2 Gluck, S., Cannon. C. and AI-Awqati, Q., Excocytosisregulates H* transport in the turtle bladder by rapid insertion of H ÷ pumps into the luminal membrane, Proc. Natl. Acad. Sci. U.S.A., 79 (1982) 4327-4331. 3 Hacket, J.T., Selective antagonism of frog cerebeilar synaptic transmission by manganese and cobalt ions, Brain Re-
transient in the synaptic gap and its vicinity. The other possibility is that synaptic vesicles insert their proton ATPases into the nerve terminal on fusion. These proton pumps can now 'turn over' and reduce the synaptic cleft pH. Similar studies in epithelia have shown that exocytosis of acid vesicles causes insertion of proton pumps into the membrane with a consequent increase in proton transport across the membrane 1'2. Exocytotic fusion of synaptic vesicles is restricted to defined domains of the terminal. In such a case the described changes in pH should occur in the immediate vicinity of the synaptic junction and their local amplitudes should be much greater than the detected values. The cause of the delayed alkalinization requires further study. It has been shown that a considerable part of isolated neurones from the rat brain possess excitatory desensitizing responses activated by rapid drops in external pH ~4. Properties of the proton-activated Na conductance mechanism in these neurones are similar to those previously described for the rat sensory, neurones 6. The functional significance of such 'proton receptor" is so far unclear (see also ref. 7). But since this mechanism is sensitive to rapid extraceilular pH changes, it can produce depolarization of the neuronal membrane in response to the transient pH drops described in this investigation and thus contribute to the interneuronai communications. Irrespective of the validity of this assumption, the complex sequence of rapid and slow pH changes accompanying synaptic transmission is a new factor which may be relevant and important for the function of the brain. ACKNOWLEDGEMENTS The authors are grateful to Prof. Q. AI-Awqati for valuable comments on the manuscript.
search, 114 (1976) 47-52. 4 Kerkut, G.A. and Wheal, H.V., (Eds.), Electrophysiology of Isolated blammalian CNS Preparation, Academic, London, 1981. 5 Kraig, P.R., Ferreira-Fiiho, C.R. and Nicholson, C., Alkaline and acid transients in cerebellar microenvironment, J. Neurophysiol., 49 (1983) 831-850. 6 Krishtal, O.A. and Pidoplichko, V.I., A receptor for protons in the nerve cell membrane, Neuroscience, 5 (1980) 2325-2327. 7 Krishta!, O.A. and Pidioplichko. V.I., A receptor for proton in the membrane of sensory neurones may participate in
356 nociception, Neuroscience, 6 ( 1981) 2599-2601. 8 Melnik, V.I. and Glebow, R.N., Kryhanovski, G.N., ATPdependent translocation of protons across the membrane of rat brain synaptic vesicles, Bull. Exp. Biol. Med., 99 (1985) 35-38. 9 Nicholson, J. and Honnsquard, N., Diffusion in the slice microenvironment for physiological studies, Fed. Proc., 42 (1983) 2865-2868. 10 Passow, H., Fasold, H., Jennings, M.L. and Lepke, S., The study of anion transport protein ('Band 3 Protein') in the rat cell membrane by means of tritiated DIDS. In J.A. Zadunaisky (Ed.), Cloride Transport in Biological Membranes, Academic, New York, I982, pp. 1-31.
11 Piwnica-Worms, D. and Lieberman, M., Microflourometric monitoring of pH l in cultured heart cells: Na+-H ÷ exchange, Am. J. Physiol., 244 (Cell Physiol., 13) (1983) c422-428. 12 Stadler, H. and Tsukita, S., Synaptic vesicles contain an ATP-dependent proton pump and show 'knob-like' protrusions on their surface, EMBO J., 3 (1984) 3333-3337. 13 Voronin, I., Long-term potentiation in the hippocampus slices, Neuroscience, 10 (1983) 1051-1069. 14 Vrublevski, S.V., Krishtal, O.A., Osipchuk, Yu.V. and Tsyndrenko, A.Ya., Proton-activated Na-conduction in cerebral cortex neurones, Dokl. Akad. Nauk SSSR, 284 (1985) 990-992.