- Email: [email protected]
In vivo whole-cell recording from neurons of the superior colliculus in fetal rats
Developmental Brain Research 108 Ž1998. 255–262
Research report
In vivo whole-cell recording from neurons of the superior colliculus in fetal rats Yoshiyuki Sakata ) , Takashi Fujioka, Shoji Nakamura Department of Physiology, Yamaguchi UniÕersity, School of Medicine, Ube, Yamaguchi 755, Japan Accepted 17 March 1998
Abstract In vivo whole-cell recording was made from neurons of the superior colliculus ŽSC. in rat fetuses which were connected with the dam by the umbilical cord. Fast action potentials could be generated by changing membrane potentials to depolarizing direction. The firing thresholds of fetal SC neurons appeared to be higher Žbetween y50 and y35 mV. than those of brain neurons in later developmental stages. The action potentials of fetal SC neurons, which were smaller in amplitude and wider in duration as compared to mature brain neurons, revealed a small or no afterhyperpolarization. In addition to these fast action potentials, slow depolarizations of smaller amplitudes were evoked by intracellular injection of long depolarization pulses. Most neurons of the fetal SC Ž7r11. revealed linear current–voltage Ž I–V . relations, while the remaining neurons displayed marked rectification. In some fetal SC neurons, presumed EPSPs and IPSPs occurred spontaneously. These presumed postsynaptic potentials showed temporal summation. These results suggest that rat SC neurons are functionally active even before birth, though the membrane properties remain immature. q 1998 Elsevier Science B.V. All rights reserved. Keywords: Membrane property; Development; Action potential; Rectification; Postsynaptic potential
1. Introduction In the developing brain, neuronal activity is known to play an important role in the normal establishment of the neuronal connection between nerve cells through unknown mechanisms w10,27,35x. In order to understand the mechanisms of the trophic action of neuronal activity, it is essential to accumulate knowledge on the electrical activity of brain neurons during development. Regarding in vitro experiments, data on the neuronal activity of brain neurons in early developmental stages has become available. However, only few investigations have been concerned with in vivo electrical activity in the developing brain, because of technical difficulties. Recently, we have developed a simple method for recording the in vivo electrical activity of brain neurons of rats during early developmental stages w14,23,24,29x. Using our method, extracellular recordings of action potentials of individual neurons become possible even in rat fetuses which are still connected with their dams by the ) Corresponding author. Fax: [email protected]
q 81-836-22-2211;
E-mail:
0165-3806r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved. PII S 0 1 6 5 - 3 8 0 6 Ž 9 8 . 0 0 0 5 5 - 8
intact umbilical cord. In this preparation, we have found that most brainstem neurons of rat fetuses have no spontaneous activity while the sensitivity of the neurons to the excitatory amino acid glutamate is extremely high w24x. In the present experiments, we further extended these extracellular studies to whole-cell recordings with patch electrodes to acquire knowledge on in vivo membrane properties of fetal brain neurons. For this study, we chose the superior colliculus ŽSC. as a recording site for the following reasons: Ž1. the development of morphology of the retino-collicular system has extensively been studied and data on the morphological development of the fetal SC has accumulated w5,17,32x, Ž2. in vivo electrical activity of retinal ganglion cells which send main afferent input to the SC has successfully been recorded in fetal rats w7,18x, Ž3. since the surface of the SC in fetal rats is not covered with the cerebral cortex but exposed, the extent of moving the tip of a recording electrode can be limited to a short distance Žup to 600 m m. from the surface of the SC. This is advantageous to whole-cell recordings in which the tip of the patch electrode should be kept clean. This report describes for the first time in vivo membrane properties of mammalian brain neurons before birth.
256
Y. Sakata et al.r DeÕelopmental Brain Research 108 (1998) 255–262
Portions of the present results have been previously reported in abstract form w30x. 2. Materials and methods The pregnant rats ŽSprague–Dawley rats, 3 months old. were housed in separate cages during gestation beginning with the first day on which they were sperm positive Žembryonic day 1, E1.. The number of pregnant rats used was 46. In the present experiments, fetuses at E20–22 were used. The pregnant rats were anesthetized with urethan Ž1.2–1.4 grkg, i.p.. and supplemented Ž0.2–0.3 grkg, i.p.. as necessary during experiment w29x. If required, the anesthetic Ž0.01 mgrg, i.p.. was injected directly to the fetus. The fetus was paralyzed with gallamine triethiodide Ž0.01 mgrg, i.p... One or two fetuses were used for each pregnant rat. Body temperature of dams was maintained at 37 " 18C by a heating pad. For fetal rats, warm saline Ž; 378C. was frequently poured on the surface of the skin and uterus to maintain an appropriate body temperature and humid environment, with weak radiation of infrared lamp Ž25 W. at 30 cm above the fetus. The method for head fixation of fetal rats to a conventional stereotaxic apparatus has been described elsewhere w23,24,29x. Briefly, pregnant rats were laid in an open box made of acrylic boards with a heating pad on the bottom. The uterus was exposed by cesarean section and partially cut to exposure of the fetus. The fetal body was placed in a syringe tube Ždiameter, about 18 mm. which was fixed on a stainless steel board by an adhesive tape. The umbilical cord was passed through a small opening at the bottom of the syringe tube and covered with a cotton. The fetal head was attached to a stereotaxic apparatus with a simple device made with a small stainless steel tube. After the skin over the cranium was removed, the rostral part of the skull was glued directly to the wire with dental cement and cyanoacrylate glue. Care was taken to ensure that the skull was horizontal between bregma and lambda and the head was not tilted to one side. In vivo whole-cell recording was applied by slightly modifying the methods described previously w1,3,4,25x. Electrodes were pulled in one stage from thin-wall borosilicate glass ŽClark Electromedical Instruments, GC150TF10, o.d. 1.5 mm, with a filament. on a vertical puller ŽNarishige, PE-2.. These had tip i.d. between 1.5–2.0 m m and resistances of 4–7 M V. The electrodes were filled with the following solution w3x Žin mM.: K-gluconate, 130; CaCl 2 , 1; MgCl 2 , 1; ethylenebisŽoxonitrilo.-tetraacetate ŽEGTA ., 1.1; 4-Ž2-hydroxyethyl.-1-piperazineethanesulphonic acid ŽHEPES., 10; and K 2 ATP, 2; pH 7.35 adjusted by KOH. The electrode solution was filtered through millipore filter Ž0.22 m m., before it was filled in the electrodes. The electrode was inserted vertically into the superficial and deep layer of the SC Žthe depth from ventral surface, 0–600 m m. by a hydraulic manipulator ŽNarishige, MO10..
A small positive pressure Ž30–40 mmHg. was applied to keep the tip of the electrode clean. The electrode holder was connected to a 2-ml glass syringe by silicon tubing filled with mineral oil to set on a micro-manipulator ŽNarishige, SM10. and the pressure of the inside of the electrode was monitored with a pressure transducer ŽNihon Kohden, DX360.. A patch-clamp amplifier ŽAxon Instruments, Axopatch-1D. was used for recording intracellular membrane potential with current clamp mode. Neurons were explored by advancing the electrode into the SC, monitoring amplified extracellular signals and a seal resistance while a current-step pulse Žy0.1 nA, 500 ms. was applied. The seal resistance Ž) 200 M V . became high when the electrode tip was placed close to a cell membrane. A small negative pressure Žy20 ; y30 mmHg. was then applied through the syringe. At the same time, formation of a giga-Ohm seal Ž) 1 GV . was monitored. When the electrode tip was relatively clean, a giga-Ohm seal was easily achieved by a low negative pressure Žy10 mmHg.. After maintaining a giga-Ohm seal at 0 mmHg, 40 mmHg negative pressure was again applied to rupture the patch membrane. In addition to this negative pressure, a single shot hyperpolarizing pulse Žy0.6 nA, 30 ms. was given to obtain the rupture of the membrane. When whole cell recording was obtained, negative pressure was immediately returned to 0 mmHg. The series resistance Ž20–50 M V . and capacitance of the patch electrode were compensated. Junction potentials might result from the Donnan potential across the perforated membrane, caused by impermeative intracellular anions w9x. Assuming concentration of the impermeative monovalent anion from 100 to 150 mM, the Donnan potential was estimated to be q8.6 to y1.8 mV w26x. However, values of membrane potentials were not corrected for junction potentials, and membrane potentials measured were within the range of those obtained by conventional whole-cell recordings. One electrode was used to search for neurons along only one track. The membrane properties of fetal SC neurons were measured as follows. A spike amplitude was obtained by subtracting holding potential Ž V h . from the peak value of the first action potential evoked by a depolarizing current pulse Ž20–40 pA.. The rise time of the action potential was time to rise from 10% of peak to 90%. The decay time was time to fall from 90% of peak to 10%. The duration of the spike was measured at the upper one-third of the Vh-to-peak amplitude. Membrane input resistance Ž R i . was estimated from the linear portion of steady-state current–voltage Ž I–V . relation obtained near the end of voltage responses to hyperpolarizing current. A membrane time constant was determined by fitting a single exponential function ŽBoltzmann’s equation. to the onset of voltage responses evoked with hyperpolarizing current Žy30 pA.. Data were stored on magnetic tape in a data-recorder ŽTEAC, RD-135. and computer hard disc ŽTISHO, VOW486M1r66.. All the data were analyzed by pCLAMP6 software ŽAxon Instruments..
Y. Sakata et al.r DeÕelopmental Brain Research 108 (1998) 255–262
257
Fig. 1. Action potentials of a fetal SC neuron. ŽA. Action potentials of a fetal SC neuron at different holding potentials Ž Vh .. The top and second traces show action potentials generating at Vh of y27 mV in fast Žtop. and slow Žsecond. chart speed of the recorder. The third trace is a record of action potentials at V h of y33 mV. The discharge rate was markedly reduced and the spike amplitude was increased by approximately 38%, when the Vh was changed from y27 to y33 mV. ŽB. Action potentials of the same neuron evoked by intracellular injection of depolarizing current pulses Ž20, 30, 40 pA, 400 ms duration.. The number of spikes evoked was increased, as the stimulus currents were increased.
Fig. 2. Responses of fetal SC neurons to intracellular injection of depolarizing current pulses Ž20, 30, 40 pA, 400 ms duration.. ŽA. This fetal SC neuron revealed a slow action potential and slow depolarization Žindicated by arrows., as well as fast action potentials in response to depolarizing current pulses. Note that an increase in the stimulus intensity led to an inactivation of fast action potentials. ŽB. In this neuron, the number of fast action potentials increased as the stimulus intensity was increased.
258
Resting potential ŽmV.
Ri ŽM V .
Time constant Žms.
Action potential Amplitude ŽmV.
Rise time Žms.
Decay time Žms.
Half duration Žms.
y40 Ž2. Ž ns 29. y60 y45
650 Ž90. Ž ns11. 110 52
22.0 Ž3.7. Ž ns11. 50 4.2
61 Ž3. Ž ns 25. 60 y
13.0 Ž2.5. Ž ns 25. y y
14.2 Ž2.1. Ž ns 25. y y
13.0 Ž2.1. a Ž ns 25. 16 y
y72 y75.3 y59.4
120 24.3 26.6
28 8.2 4.1
120 y 78.7
y y y
y y y
2 y 1–1.5
y67.4
116
y
y
y
y
y
Values are means Ž"S.E... See Section 2 in the text about the definition of the membrane properties. The duration of the spike was measured at the upper one-third of the Vh -to-peak amplitude.
a
superior colliculus E20–22 Žrat. in vivo hippocampus PD3 Žrat. slice, Spigelman et al. w33x neostriatum Žrat. in vivo PD6–10 Žrat. in vivo, Tepper and Trend w34x hippocampus PD30 Žrat. slice, Spigelman et al. w33x neocortex adult Žrat. slice, Kriegstein et al. w15x superior colliculus adult Žguinea pig. slice, Lopez-Barned and Llinas w16x superior colliculus adult Žhamster. slice, Mooney et al. w22x
Y. Sakata et al.r DeÕelopmental Brain Research 108 (1998) 255–262
Table 1 In vivo membrane properties of fetal SC neurons in comparison with those of in vitro hippocampal, neocortical and in vivo neostriatal neurons at different developmental stages
Y. Sakata et al.r DeÕelopmental Brain Research 108 (1998) 255–262
3. Results In vivo whole-cell recordings were successful in a total of 29 fetal SC neurons. Action potentials and resting membrane properties of these neurons were analyzed. 3.1. Action potentials Action potentials could be generated by depolarizing the cell. At holding potentials Ž V h . more negative than y50 mV no action potentials occurred, while V h more positive than y35 mV generated action potentials Ž n s 10. ŽFig. 1.. Thus firing thresholds of fetal SC neurons appear to be higher Žbetween y50 and y35 mV. than those of brain neurons in later developmental stages Žy55 mV., such as spinal moto-neurons in adult cats w13x. When Vh was changed from depolarized to hyperpolarized potentials, the firing frequency of action potentials decreased and the amplitudes became larger ŽFig. 1A.. Action potentials of fetal SC neurons could be evoked by intracellular injection of depolarizing pulses at Vh below the subthresholds of spike firing Žy50 to y60 mV. ŽFig. 2.. Because depolarizing pulses with short duration Ž10–30 ms. were difficult to evoke action potentials, pulses
259
of long duration Ž400 ms. were applied. Single or multiple action potentials were generated with a long delay after the onset of depolarizing pulses, while increasing stimulus currents resulted in reducing latencies of the first action potentials ŽFig. 1B, Fig. 2.. Some cells Ž6r11. revealed an inactivation of action potentials ŽFig. 2A. and in the remaining cells Ž n s 5. the number of the action potentials increased ŽFig. 2B., as the intracellular depolarizing currents were increased Ž20, 30, 40 pA.. The action potentials of fetal SC neurons evoked by intracellular injection of depolarizing pulses were different in several aspects from those of mature brain neurons. The action potentials of fetal SC neurons were smaller in amplitude and wider in duration as compared to mature brain neurons, including SC neurons in slice preparations of the adult guinea pig Žsee Table 1. w16x. The amplitudes of the action potentials occurring at resting potentials of y34 to y48 mV ranged from 30 to 49 mV Žmean " S.E., 37 " 4 mV, n s 5.. The amplitudes of the first action potentials evoked by intracellular depolarizing pulses at Vh of y50 to y73 mV ranged from 35 to 93 mV Ž61 " 3 mV, n s 25., and their duration from 2.8 to 40 ms Ž13.0 " 2.1 ms, n s 25., as shown in Table 1. In multiple action potentials induced by depolarization pulses, the duration of
Fig. 3. Current–voltage Ž I–V . relationship. ŽA1. Membrane responses of a fetal SC neuron to intracellular injection of hyperpolarizing and depolarizing current pulses at V h of y65 mV. The stimulus pulses of 500 ms duration with currents ranging from y40 to 40 pA were given at a step of 10 pA. This fetal SC neuron revealed the slow depolarization Žindicated by arrows., as well as fast action potentials in response to depolarizing current pulses. ŽA2. I–V plots of the neuron constructed near the end of the response shown in ŽA.. Note the nearly linear I–V relationship. ŽB1. Membrane responses to intracellular injection of hyperpolarizing and depolarizing current pulses at Vh of y60 mV. ŽB2. I–V plots of the SC neuron. Note the rectification in the depolarizing and hyperpolarizing range.
260
Y. Sakata et al.r DeÕelopmental Brain Research 108 (1998) 255–262
the succeeding action potentials was longer than that of the first action potentials. The action potentials of fetal SC neurons revealed small or no afterhyperpolarization. In addition to the fast action potentials, slow depolarizations of smaller amplitudes Žindicated by arrows in Fig. 2A, Fig. 3A1. were evoked by intracellular injection of long depolarization pulses. These depolarizations appeared when the membrane potentials were depolarized to more than y10 mV.
3.2. Current–Õoltage relationship To further analyze the membrane properties of fetal SC neurons, I–V plots were made based on the responses of fetal SC neurons to depolarizing and hyperpolarizing current step pulses. At V h of y60 to y65 mV, depolarizing and hyperpolarizing currents ranging from y40 to 60 pA were injected directly into the cells, and changes in the membrane potentials near the end of current pulse were measured to construct the I–V plots ŽFig. 3.. In 11 fetal SC neurons examined, seven neurons revealed linear I–V relations, while the remaining four displayed marked rectifications. Three of the neurons showing non-linear I–V relations displayed anomalous rectifications, while the remaining one revealed outward rectification. The input resistance Ž R i . of fetal SC neurons estimated from a slope of the regression line of I–V relations ranged from 234 to 1168 M V Ž650 " 90 M V, n s 11.. The mean membrane time constant was 22.0 msŽS.E., 3.7 ms, n s 11.. Regarding the resting potential, mean R i and membrane time constant, there was no significant difference between the two groups of the fetal SC neurons showing the linear and non-linear I–V relations Žresting potential: linear, 39 " 3 mV, n s 7; non-linear, 37 " 3 mV, n s 4, R i : linear, 733 " 117 M V, n s 7; non-linear, 555 " 175 M V, n s 4, membrane time constant: linear, 24.7 " 5.0 ms, n s 7; non-linear, 17.6 " 4.7 ms, n s 4..
3.3. Presumed postsynaptic potential In the present experiments, some fetal SC neurons revealed small potentials similar to EPSPs Žtwo neurons. and IPSPs Žthree neurons. during recordings, although further experiments need to assess the involvement of synaptic transmission in these potentials. These presumed postsynaptic potentials occurred spontaneously. In one of these neurons, both presumed EPSPs and IPSPs appeared at V h of y55 mV ŽFig. 4A.. Temporal summation of these depolarizing and hyperpolarizing potentials was observed in addition to monophasic potentials. The presumed EPSPs were 1.0 to 2.8 mV Žmonophasic, 1.3 " 0.2 mV, n s 4. in amplitude and 74 to 93 ms Ž81.0 " 4.3 ms, n s 4. in
Fig. 4. Presumed EPSPs and IPSPs in fetal SC neurons. ŽA. This neuron revealed both presumed EPSPs and IPSPs Žarrows. at V h of y55 mV. Some EPSPs showed temporal summation. ŽB. In this neuron, only IPSPs Žarrows. were observed at V h of y55 mV. One IPSP showed temporal summation. The upper records of ŽA. and ŽB. indicate action potentials evoked by a depolarizing current pulse Ž40 pA..
duration. The amplitude and duration of presumed IPSPs Žmonophasic IPSPs. were 4.0 mV ŽS.E.,0.8 mV, n s 6. and 124.0 ms ŽS.E.,18.0 ms, n s 6., respectively.
4. Discussion This paper describes for the first time in vivo membrane properties of brain neurons of the fetal rat, which is still connected with the dam by the intact umbilical cord. The membrane properties of the fetal SC neurons revealed immaturity in action potentials and resting membrane properties. Although there are no comparable data on the membrane properties of fetal and neonatal SC neurons, Spigelman et al. w33x have reported whole-cell recordings of hippocampal neurons in brain slices obtained from rats of postnatal day ŽPD. 2 to 30. The membrane properties of hippocampal neurons of PD2–5 in this in vitro study resemble those of fetal SC neurons recorded in the present experiments. In the neonatal hippocampal neurons, the mean amplitude of action potentials was approximately 60 mV and the mean duration was about 16 ms ŽTable 1.. They also reported the large R i Ž1100 M V . and long time constant Ž50 ms. for the neonatal hippocampal neurons ŽTable 1.. The resting membrane potential of these immature neurons Žmean s y60 mV. was more depolarized than that of mature neurons Že.g., y72 mV in sliced
Y. Sakata et al.r DeÕelopmental Brain Research 108 (1998) 255–262
hippocampal neurons at PD30. ŽTable 1. w33x. Compared to the neonatal hippocampal neurons, the membrane properties of the fetal SC neurons appear to be more immature. Smaller spike amplitude and longer spike duration in immature brain neurons have been also reported in other brain regions ŽTable 1. w11,15,33,34x. The resting membrane potentials andror the kinetics and density of ion channels underlying spike genesis and repolarization are the likely causes of the smaller spike amplitudes and longer spike durations observed in the present experiments. It is unlikely that the membrane potential is a cause of the small spike amplitudes and longer durations, because they were measured by holding a membrane potential near y60 mV. Thus the kinetics and density of ion channels of fetal SC neurons are thought to be different from those of more matured neurons ŽTable 1. w16,33x. Previous studies have indicated that the smaller spike amplitude of immature brain neurons is due to a low density of functional sodium channels w11,33x. It is likely that the same mechanism underlies the smaller spike amplitude of fetal SC neurons. The lower density of sodium channels may also account for the higher threshold for spike generation observed in fetal SC neurons Žbetween y50 and y35 mV.. In the present experiments, however, it remains to be determined if either the total number of sodium channels or the number of functional sodium channels on fetal SC neurons is smaller as compared to more matured neurons. Since most fetal SC neurons recorded did not reveal afterhyperpolarization, potassium channels involved in the genesis of afterhyperpolarization are considered to be not fully developed in these immature neurons. The lack of afterhyperpolarization may contribute partly to the longer spike duration of fetal SC neurons. In addition, the immaturity of sodium–potassium pumps may also be attributable to the spike configuration of the immature neurons w6x. The slow depolarizations of smaller amplitudes were often observed in fetal SC neurons. These depolarizations with a slow rising and falling phase did not occur at the resting membrane potentials but were evoked by intracellular injection of a long-duration depolarizing pulse. Furthermore, this depolarization occurred with a long delay, when the multiple spikes were induced by a long depolarizing pulse. These findings indicate that the spikelike depolarization is evoked under the state of a long-lasting depolarization. Thus it is likely that the observed depolarizations result from a partial inactivation of sodium channels andror are associated with calcium influx. Immature brain neurons are known to reveal more linear I–V relations than adult neurons, although rectification was often present in both the hyperpolarizing and depolarizing range w19x. Similar features of I–V relations were also observed in fetal SC neurons. The majority Ž7r11. of the fetal SC neurons examined exhibited no anomalous rectification. This finding is consistent with the
261
in vitro studies that reported the absence of anomalous rectifications in hippocampal w31x and neostriatal neurons w20x at early developmental stages. Furthermore, in a developmental study on rat neostriatal neurons using in vivo preparations, Tepper and Trend w34x have demonstrated that in rats younger than PD11, only 13% of neostriatal neurons examined displayed anomalous rectifications and the proportion of neostriatal neurons showing anomalous rectifications increased with development. On the other hand, a small fraction Ž20%. of neostriatal neurons from the first and second postnatal weeks exhibits the outward rectification w34x. Itaya et al. w12x and Molotchnikoff and Itaya w21x reported that the spontaneous activity of SC neurons in rats first appeared at PD5, while electrical stimulation of the optic nerve could not produce evoked response in the SC until PD10 but increased the spontaneous discharge of neonatal SC neurons. Based on these findings, they suggested that despite the presence of spontaneous activity of retinal ganglion cells w18x, classical synaptic transmission did not occur in the rat SC during prenatal periods. In contrast, Reece and Lim reported that in in vitro preparations, optic nerve stimulation could elicit evoked responses in the SC as early as E18 w28x. These responses were blocked by a calcium blocker and a glutamatergic antagonist thus suggesting their postsynaptic origin. Although the discrepancy between the in vivo and the in vitro study remains to be explained, the present experiments, in which presumed postsynaptic potentials were recorded in fetal SC neurons, may support the presence of synaptic transmission in the SC before birth. Furthermore, the higher threshold for spike generation of fetal SC neurons observed in the present experiments may explain the failure of Itaya et al. to obtain the SC responses of the immature brain to optic nerve stimulation. Using the patch clamp recording, Grantyn et al. w8x have found that 44 h after the start of dissociated cell cultures derived from the SC of E21 rats, spontaneous bicuculline-sensitive Cly currents could be recorded. The frequency of the currents could be increased by superfusing the cultured cells with kainate and reduced by addition of blockers of voltage-dependent Ca2q channels. This study suggests that spontaneous IPSPs mediated by GABA occur in the SC at early developmental stages, probably before birth. 5. Unlinked References w2x Acknowledgements The authors are grateful to Dr. H. Onimaru and Dr. T. Shosaku for kind advices on in vivo whole-cell recording technique and for reviewing the manuscript. This work was partly supported by a Grant-in-Aid for Science Re-
262
Y. Sakata et al.r DeÕelopmental Brain Research 108 (1998) 255–262
search ŽNo. 06454475. from the Ministry of Education, Science and Culture of Japan.
References w1x M.G. Blanton, J.J. Lo Turco, A.R. Kriegstein, Whole cell recording from neurons in slices of reptilian and mammalian cerebral cortex, J. Neurosci. Methods 30 Ž1989. 203–210. w2x O.S. Cardoso, K.P. Hoffman, The cortico-tectal projection of the rat in vitro: development, anatomy and physiological characteristics, Eur. J. Neurosci. 7 Ž1995. 599–612. w3x J.B. Dean, E.A. Gallman, D.E. Millhorn, Whole-cell recordings in the dorsal vagal complex ŽDVC. in thick slices prepared from adult rat, in: D.T. Frazier, M.S. Dekin, W.R. Revelette, D.F. Speck ŽEds.., International Conference on Modulation of Respiratory Pattern: Peripheral and Central Mechanisms, Lexington, KY, USA, 1990, pp. 57. w4x F.A. Edwards, A. Konnerth, B. Sakmann, T. Takahashi, A thin slice preparation for patch clamp recordings from neurons of the mammalian central nervous system, Pflug. ¨ Arch. Eur. J. Physiol. 414 Ž1989. 600–612. w5x R.E. Foster, B.W. Conners, S.G. Waxman, Rat optic nerve: electrophysiological, pharmacological and anatomical studies during development, Dev. Brain Res. 3 Ž1982. 371–386. w6x A. Fukuda, D.A. Prince, Postnatal development of electrogenic sodium pump activity in rat hippocampal pyramidal neurons, Dev. Brain Res. 65 Ž1992. 101–114. w7x L. Galli, L. Maffei, Spontaneous impulse activity of rat retinal ganglion cells in prenatal life, Science 245 Ž1988. 90–91. w8x R. Grantyn, K. Krazewski, I. Melnick, H. Taschenberger, S.S. Warton, In vitro development of vertebrate central synapses, Perspect. Dev. Neurol. 2 Ž1995. 387–397. w9x R. Horn, A. Marty, Muscarinic activation of ionic currents measured by a new whole cell recording method, J. Gen. Physiol. 92 Ž1988. 145–159. w10x D.H. Hubel, T.N. Wiesel, Binocular interaction in striate cortex of kittens reared with artificial squint, J. Neurophysiol. 28 Ž1965. 1041–1059. w11x J.R. Huguenard, O.P. Hamill, D.A. Prince, Developmental changes in Naq conductances in rat neocortical neurons: appearance of a slowly inactivating component, J. Neurophysiol. 59 Ž1988. 778–779. w12x S.K. Itaya, S. Fortin, S. Molotchnikoff, Evolution of spontaneous activity in the developing rat superior colliculus, Can. J. Physiol. Pharmacol. 73 Ž1995. 1372–1377. w13x E.R. Kandel, J.H. Schwartz, Directly gated transmission at the nerve–muscle synapse, in: E.R. Kandel, J.H. Schwartz, T.M. Jessell ŽEds.., Principles of Neural Science, Elsevier, Tokyo, 1991, pp. 153–172. w14x F. Kimura, S. Nakamura, Locus coeruleus neurons in the neonatal rats: electrical activity and responses to sensory stimulation, Dev. Brain Res. 23 Ž1985. 301–305. w15x A.R. Kriegstein, T. Suppes, D.A. Prince, Cellular and synaptic physiology and epileptogenesis of developing rat neocortical neurons in vitro, Dev. Brain Res. 34 Ž1986. 161–171.
w16x J. Lopez-Barned, R. Llinas, Electrophysiology of mammalian tectal neurons in vitro: I. Transient ionic conductance, J. Neurophysiol. 60 Ž1988. 853–868. w17x R.D. Lund, A.H. Bunt, Prenatal development of central optic pathways in albino rats, J. Comp. Neurol. 167 Ž1976. 247–264. w18x L. Maffei, L. Galli-Rosta, Correlation in the discharges of neighboring rat retinal ganglion cells during prenatal life, Proc. Natl. Acad. Sci. U.S.A. 87 Ž1990. 2861–2864. w19x D.A. McCormick, D.A. Prince, Postnatal development of electrophysiological properties of rat cerebral cortical pyramidal neurons, J. Physiol. 393 Ž1987. 743–762. w20x U. Misgeld, H.U. Dodt, M. Frotscher, Late development of intrinsic excitation in the rat neostriatum: an in vitro study, Dev. Brain Res. 27 Ž1986. 59–67. w21x S. Molotchnikoff, S.K. Itaya, Functional development of the neonatal retino-tectal pathway, Dev. Brain Res. 72 Ž1993. 300–304. w22x R.D. Mooney, M.Y. Shi, R.W. Rhoades, Modulation of retino-tectal transmission by presynaptic 5-HT1B receptors in the superior colliculus of the adult hamster, J. Neurophysiol. 72 Ž1994. 3–13. w23x S. Nakamura, T. Sakaguchi, Development and plasticity of the locus coeruleus: a review of recent physiological and pharmacological experimentation, Prog. Neurobiol. 34 Ž1990. 505–526. w24x S. Nakamura, S. Arakawa, S. Nishiike, A long-lasting change in the excitability of fetal brain neurons following activation of dam’s hypothalamus in rats, Am. J. Physiol. 269 Ž1995. R236–R244. w25x H. Onimaru, I. Homma, Whole cell recordings from respiratory neurons in the medulla of brainstem–spinal cord preparations isolated from newborn rat, Pflug. ¨ Arch. Eur. J. Physiol. 420 Ž1992. 399–406. w26x H. Onimaru, A. Arata, I. Homma, Intrinsic burst generation of preinspiratory neurons in the medulla of brainstem–spinal cord preparations isolated from newborn rats, Exp. Brain Res. 106 Ž1995. 57–68. w27x D. Purves ŽEd.., Body and Brain, Harvard Univ. Press, Cambridge, MA, 1988. w28x L.J. Reece, C.H. Lim, Optic nerve stimulation elicits evoked responses in superior colliculus of embryonic rats in vitro, Neurosci. Abstr. 21 Ž1995. 1503. w29x T. Sakaguchi, S. Nakamura, Some in vivo electrophysiological properties of locus coeruleus neurons in fetal rats, Exp. Brain Res. 68 Ž1987. 122–130. w30x Y. Sakata, T. Fujioka, Y. Fujii, I. Akimura, S. Nakamura, In vivo whole-cell recordings of superior collicular neurons in fetal rats, Neurosci. Res. 20 Ž1996. S124, Suppl. w31x M. Segal, Developmental changes in serotonin actions in rat hippocampus, Dev. Brain Res. 52 Ž1990. 247–252. w32x D.K. Simon, D.D.M. O’Leary, Responses of retinal axons in vivo and in vitro to position-encoding molecules in the embryonic superior colliculus, Neuron 9 Ž1992. 977–989. w33x I. Spigelman, L. Zhang, P.L. Carlen, Patch-clamp study of postnatal development of CA1 neurons in rat hippocampal slices: membrane excitability and Kq currents, J. Neurophysiol. 68 Ž1992. 55–69. w34x J.M. Tepper, F. Trend, In vivo studies of the postnatal development of rat neostriatal neurons, Prog. Brain Res. 99 Ž1992. 35–50. w35x T.N. Wiesel, D.H. Hubel, Effects of visual deprivation on morphology and physiology of cells in the cat’s lateral geniculate body, J. Neurophysiol. 26 Ž1963. 978–993.
Research report
In vivo whole-cell recording from neurons of the superior colliculus in fetal rats Yoshiyuki Sakata ) , Takashi Fujioka, Shoji Nakamura Department of Physiology, Yamaguchi UniÕersity, School of Medicine, Ube, Yamaguchi 755, Japan Accepted 17 March 1998
Abstract In vivo whole-cell recording was made from neurons of the superior colliculus ŽSC. in rat fetuses which were connected with the dam by the umbilical cord. Fast action potentials could be generated by changing membrane potentials to depolarizing direction. The firing thresholds of fetal SC neurons appeared to be higher Žbetween y50 and y35 mV. than those of brain neurons in later developmental stages. The action potentials of fetal SC neurons, which were smaller in amplitude and wider in duration as compared to mature brain neurons, revealed a small or no afterhyperpolarization. In addition to these fast action potentials, slow depolarizations of smaller amplitudes were evoked by intracellular injection of long depolarization pulses. Most neurons of the fetal SC Ž7r11. revealed linear current–voltage Ž I–V . relations, while the remaining neurons displayed marked rectification. In some fetal SC neurons, presumed EPSPs and IPSPs occurred spontaneously. These presumed postsynaptic potentials showed temporal summation. These results suggest that rat SC neurons are functionally active even before birth, though the membrane properties remain immature. q 1998 Elsevier Science B.V. All rights reserved. Keywords: Membrane property; Development; Action potential; Rectification; Postsynaptic potential
1. Introduction In the developing brain, neuronal activity is known to play an important role in the normal establishment of the neuronal connection between nerve cells through unknown mechanisms w10,27,35x. In order to understand the mechanisms of the trophic action of neuronal activity, it is essential to accumulate knowledge on the electrical activity of brain neurons during development. Regarding in vitro experiments, data on the neuronal activity of brain neurons in early developmental stages has become available. However, only few investigations have been concerned with in vivo electrical activity in the developing brain, because of technical difficulties. Recently, we have developed a simple method for recording the in vivo electrical activity of brain neurons of rats during early developmental stages w14,23,24,29x. Using our method, extracellular recordings of action potentials of individual neurons become possible even in rat fetuses which are still connected with their dams by the ) Corresponding author. Fax: [email protected]
q 81-836-22-2211;
E-mail:
0165-3806r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved. PII S 0 1 6 5 - 3 8 0 6 Ž 9 8 . 0 0 0 5 5 - 8
intact umbilical cord. In this preparation, we have found that most brainstem neurons of rat fetuses have no spontaneous activity while the sensitivity of the neurons to the excitatory amino acid glutamate is extremely high w24x. In the present experiments, we further extended these extracellular studies to whole-cell recordings with patch electrodes to acquire knowledge on in vivo membrane properties of fetal brain neurons. For this study, we chose the superior colliculus ŽSC. as a recording site for the following reasons: Ž1. the development of morphology of the retino-collicular system has extensively been studied and data on the morphological development of the fetal SC has accumulated w5,17,32x, Ž2. in vivo electrical activity of retinal ganglion cells which send main afferent input to the SC has successfully been recorded in fetal rats w7,18x, Ž3. since the surface of the SC in fetal rats is not covered with the cerebral cortex but exposed, the extent of moving the tip of a recording electrode can be limited to a short distance Žup to 600 m m. from the surface of the SC. This is advantageous to whole-cell recordings in which the tip of the patch electrode should be kept clean. This report describes for the first time in vivo membrane properties of mammalian brain neurons before birth.
256
Y. Sakata et al.r DeÕelopmental Brain Research 108 (1998) 255–262
Portions of the present results have been previously reported in abstract form w30x. 2. Materials and methods The pregnant rats ŽSprague–Dawley rats, 3 months old. were housed in separate cages during gestation beginning with the first day on which they were sperm positive Žembryonic day 1, E1.. The number of pregnant rats used was 46. In the present experiments, fetuses at E20–22 were used. The pregnant rats were anesthetized with urethan Ž1.2–1.4 grkg, i.p.. and supplemented Ž0.2–0.3 grkg, i.p.. as necessary during experiment w29x. If required, the anesthetic Ž0.01 mgrg, i.p.. was injected directly to the fetus. The fetus was paralyzed with gallamine triethiodide Ž0.01 mgrg, i.p... One or two fetuses were used for each pregnant rat. Body temperature of dams was maintained at 37 " 18C by a heating pad. For fetal rats, warm saline Ž; 378C. was frequently poured on the surface of the skin and uterus to maintain an appropriate body temperature and humid environment, with weak radiation of infrared lamp Ž25 W. at 30 cm above the fetus. The method for head fixation of fetal rats to a conventional stereotaxic apparatus has been described elsewhere w23,24,29x. Briefly, pregnant rats were laid in an open box made of acrylic boards with a heating pad on the bottom. The uterus was exposed by cesarean section and partially cut to exposure of the fetus. The fetal body was placed in a syringe tube Ždiameter, about 18 mm. which was fixed on a stainless steel board by an adhesive tape. The umbilical cord was passed through a small opening at the bottom of the syringe tube and covered with a cotton. The fetal head was attached to a stereotaxic apparatus with a simple device made with a small stainless steel tube. After the skin over the cranium was removed, the rostral part of the skull was glued directly to the wire with dental cement and cyanoacrylate glue. Care was taken to ensure that the skull was horizontal between bregma and lambda and the head was not tilted to one side. In vivo whole-cell recording was applied by slightly modifying the methods described previously w1,3,4,25x. Electrodes were pulled in one stage from thin-wall borosilicate glass ŽClark Electromedical Instruments, GC150TF10, o.d. 1.5 mm, with a filament. on a vertical puller ŽNarishige, PE-2.. These had tip i.d. between 1.5–2.0 m m and resistances of 4–7 M V. The electrodes were filled with the following solution w3x Žin mM.: K-gluconate, 130; CaCl 2 , 1; MgCl 2 , 1; ethylenebisŽoxonitrilo.-tetraacetate ŽEGTA ., 1.1; 4-Ž2-hydroxyethyl.-1-piperazineethanesulphonic acid ŽHEPES., 10; and K 2 ATP, 2; pH 7.35 adjusted by KOH. The electrode solution was filtered through millipore filter Ž0.22 m m., before it was filled in the electrodes. The electrode was inserted vertically into the superficial and deep layer of the SC Žthe depth from ventral surface, 0–600 m m. by a hydraulic manipulator ŽNarishige, MO10..
A small positive pressure Ž30–40 mmHg. was applied to keep the tip of the electrode clean. The electrode holder was connected to a 2-ml glass syringe by silicon tubing filled with mineral oil to set on a micro-manipulator ŽNarishige, SM10. and the pressure of the inside of the electrode was monitored with a pressure transducer ŽNihon Kohden, DX360.. A patch-clamp amplifier ŽAxon Instruments, Axopatch-1D. was used for recording intracellular membrane potential with current clamp mode. Neurons were explored by advancing the electrode into the SC, monitoring amplified extracellular signals and a seal resistance while a current-step pulse Žy0.1 nA, 500 ms. was applied. The seal resistance Ž) 200 M V . became high when the electrode tip was placed close to a cell membrane. A small negative pressure Žy20 ; y30 mmHg. was then applied through the syringe. At the same time, formation of a giga-Ohm seal Ž) 1 GV . was monitored. When the electrode tip was relatively clean, a giga-Ohm seal was easily achieved by a low negative pressure Žy10 mmHg.. After maintaining a giga-Ohm seal at 0 mmHg, 40 mmHg negative pressure was again applied to rupture the patch membrane. In addition to this negative pressure, a single shot hyperpolarizing pulse Žy0.6 nA, 30 ms. was given to obtain the rupture of the membrane. When whole cell recording was obtained, negative pressure was immediately returned to 0 mmHg. The series resistance Ž20–50 M V . and capacitance of the patch electrode were compensated. Junction potentials might result from the Donnan potential across the perforated membrane, caused by impermeative intracellular anions w9x. Assuming concentration of the impermeative monovalent anion from 100 to 150 mM, the Donnan potential was estimated to be q8.6 to y1.8 mV w26x. However, values of membrane potentials were not corrected for junction potentials, and membrane potentials measured were within the range of those obtained by conventional whole-cell recordings. One electrode was used to search for neurons along only one track. The membrane properties of fetal SC neurons were measured as follows. A spike amplitude was obtained by subtracting holding potential Ž V h . from the peak value of the first action potential evoked by a depolarizing current pulse Ž20–40 pA.. The rise time of the action potential was time to rise from 10% of peak to 90%. The decay time was time to fall from 90% of peak to 10%. The duration of the spike was measured at the upper one-third of the Vh-to-peak amplitude. Membrane input resistance Ž R i . was estimated from the linear portion of steady-state current–voltage Ž I–V . relation obtained near the end of voltage responses to hyperpolarizing current. A membrane time constant was determined by fitting a single exponential function ŽBoltzmann’s equation. to the onset of voltage responses evoked with hyperpolarizing current Žy30 pA.. Data were stored on magnetic tape in a data-recorder ŽTEAC, RD-135. and computer hard disc ŽTISHO, VOW486M1r66.. All the data were analyzed by pCLAMP6 software ŽAxon Instruments..
Y. Sakata et al.r DeÕelopmental Brain Research 108 (1998) 255–262
257
Fig. 1. Action potentials of a fetal SC neuron. ŽA. Action potentials of a fetal SC neuron at different holding potentials Ž Vh .. The top and second traces show action potentials generating at Vh of y27 mV in fast Žtop. and slow Žsecond. chart speed of the recorder. The third trace is a record of action potentials at V h of y33 mV. The discharge rate was markedly reduced and the spike amplitude was increased by approximately 38%, when the Vh was changed from y27 to y33 mV. ŽB. Action potentials of the same neuron evoked by intracellular injection of depolarizing current pulses Ž20, 30, 40 pA, 400 ms duration.. The number of spikes evoked was increased, as the stimulus currents were increased.
Fig. 2. Responses of fetal SC neurons to intracellular injection of depolarizing current pulses Ž20, 30, 40 pA, 400 ms duration.. ŽA. This fetal SC neuron revealed a slow action potential and slow depolarization Žindicated by arrows., as well as fast action potentials in response to depolarizing current pulses. Note that an increase in the stimulus intensity led to an inactivation of fast action potentials. ŽB. In this neuron, the number of fast action potentials increased as the stimulus intensity was increased.
258
Resting potential ŽmV.
Ri ŽM V .
Time constant Žms.
Action potential Amplitude ŽmV.
Rise time Žms.
Decay time Žms.
Half duration Žms.
y40 Ž2. Ž ns 29. y60 y45
650 Ž90. Ž ns11. 110 52
22.0 Ž3.7. Ž ns11. 50 4.2
61 Ž3. Ž ns 25. 60 y
13.0 Ž2.5. Ž ns 25. y y
14.2 Ž2.1. Ž ns 25. y y
13.0 Ž2.1. a Ž ns 25. 16 y
y72 y75.3 y59.4
120 24.3 26.6
28 8.2 4.1
120 y 78.7
y y y
y y y
2 y 1–1.5
y67.4
116
y
y
y
y
y
Values are means Ž"S.E... See Section 2 in the text about the definition of the membrane properties. The duration of the spike was measured at the upper one-third of the Vh -to-peak amplitude.
a
superior colliculus E20–22 Žrat. in vivo hippocampus PD3 Žrat. slice, Spigelman et al. w33x neostriatum Žrat. in vivo PD6–10 Žrat. in vivo, Tepper and Trend w34x hippocampus PD30 Žrat. slice, Spigelman et al. w33x neocortex adult Žrat. slice, Kriegstein et al. w15x superior colliculus adult Žguinea pig. slice, Lopez-Barned and Llinas w16x superior colliculus adult Žhamster. slice, Mooney et al. w22x
Y. Sakata et al.r DeÕelopmental Brain Research 108 (1998) 255–262
Table 1 In vivo membrane properties of fetal SC neurons in comparison with those of in vitro hippocampal, neocortical and in vivo neostriatal neurons at different developmental stages
Y. Sakata et al.r DeÕelopmental Brain Research 108 (1998) 255–262
3. Results In vivo whole-cell recordings were successful in a total of 29 fetal SC neurons. Action potentials and resting membrane properties of these neurons were analyzed. 3.1. Action potentials Action potentials could be generated by depolarizing the cell. At holding potentials Ž V h . more negative than y50 mV no action potentials occurred, while V h more positive than y35 mV generated action potentials Ž n s 10. ŽFig. 1.. Thus firing thresholds of fetal SC neurons appear to be higher Žbetween y50 and y35 mV. than those of brain neurons in later developmental stages Žy55 mV., such as spinal moto-neurons in adult cats w13x. When Vh was changed from depolarized to hyperpolarized potentials, the firing frequency of action potentials decreased and the amplitudes became larger ŽFig. 1A.. Action potentials of fetal SC neurons could be evoked by intracellular injection of depolarizing pulses at Vh below the subthresholds of spike firing Žy50 to y60 mV. ŽFig. 2.. Because depolarizing pulses with short duration Ž10–30 ms. were difficult to evoke action potentials, pulses
259
of long duration Ž400 ms. were applied. Single or multiple action potentials were generated with a long delay after the onset of depolarizing pulses, while increasing stimulus currents resulted in reducing latencies of the first action potentials ŽFig. 1B, Fig. 2.. Some cells Ž6r11. revealed an inactivation of action potentials ŽFig. 2A. and in the remaining cells Ž n s 5. the number of the action potentials increased ŽFig. 2B., as the intracellular depolarizing currents were increased Ž20, 30, 40 pA.. The action potentials of fetal SC neurons evoked by intracellular injection of depolarizing pulses were different in several aspects from those of mature brain neurons. The action potentials of fetal SC neurons were smaller in amplitude and wider in duration as compared to mature brain neurons, including SC neurons in slice preparations of the adult guinea pig Žsee Table 1. w16x. The amplitudes of the action potentials occurring at resting potentials of y34 to y48 mV ranged from 30 to 49 mV Žmean " S.E., 37 " 4 mV, n s 5.. The amplitudes of the first action potentials evoked by intracellular depolarizing pulses at Vh of y50 to y73 mV ranged from 35 to 93 mV Ž61 " 3 mV, n s 25., and their duration from 2.8 to 40 ms Ž13.0 " 2.1 ms, n s 25., as shown in Table 1. In multiple action potentials induced by depolarization pulses, the duration of
Fig. 3. Current–voltage Ž I–V . relationship. ŽA1. Membrane responses of a fetal SC neuron to intracellular injection of hyperpolarizing and depolarizing current pulses at V h of y65 mV. The stimulus pulses of 500 ms duration with currents ranging from y40 to 40 pA were given at a step of 10 pA. This fetal SC neuron revealed the slow depolarization Žindicated by arrows., as well as fast action potentials in response to depolarizing current pulses. ŽA2. I–V plots of the neuron constructed near the end of the response shown in ŽA.. Note the nearly linear I–V relationship. ŽB1. Membrane responses to intracellular injection of hyperpolarizing and depolarizing current pulses at Vh of y60 mV. ŽB2. I–V plots of the SC neuron. Note the rectification in the depolarizing and hyperpolarizing range.
260
Y. Sakata et al.r DeÕelopmental Brain Research 108 (1998) 255–262
the succeeding action potentials was longer than that of the first action potentials. The action potentials of fetal SC neurons revealed small or no afterhyperpolarization. In addition to the fast action potentials, slow depolarizations of smaller amplitudes Žindicated by arrows in Fig. 2A, Fig. 3A1. were evoked by intracellular injection of long depolarization pulses. These depolarizations appeared when the membrane potentials were depolarized to more than y10 mV.
3.2. Current–Õoltage relationship To further analyze the membrane properties of fetal SC neurons, I–V plots were made based on the responses of fetal SC neurons to depolarizing and hyperpolarizing current step pulses. At V h of y60 to y65 mV, depolarizing and hyperpolarizing currents ranging from y40 to 60 pA were injected directly into the cells, and changes in the membrane potentials near the end of current pulse were measured to construct the I–V plots ŽFig. 3.. In 11 fetal SC neurons examined, seven neurons revealed linear I–V relations, while the remaining four displayed marked rectifications. Three of the neurons showing non-linear I–V relations displayed anomalous rectifications, while the remaining one revealed outward rectification. The input resistance Ž R i . of fetal SC neurons estimated from a slope of the regression line of I–V relations ranged from 234 to 1168 M V Ž650 " 90 M V, n s 11.. The mean membrane time constant was 22.0 msŽS.E., 3.7 ms, n s 11.. Regarding the resting potential, mean R i and membrane time constant, there was no significant difference between the two groups of the fetal SC neurons showing the linear and non-linear I–V relations Žresting potential: linear, 39 " 3 mV, n s 7; non-linear, 37 " 3 mV, n s 4, R i : linear, 733 " 117 M V, n s 7; non-linear, 555 " 175 M V, n s 4, membrane time constant: linear, 24.7 " 5.0 ms, n s 7; non-linear, 17.6 " 4.7 ms, n s 4..
3.3. Presumed postsynaptic potential In the present experiments, some fetal SC neurons revealed small potentials similar to EPSPs Žtwo neurons. and IPSPs Žthree neurons. during recordings, although further experiments need to assess the involvement of synaptic transmission in these potentials. These presumed postsynaptic potentials occurred spontaneously. In one of these neurons, both presumed EPSPs and IPSPs appeared at V h of y55 mV ŽFig. 4A.. Temporal summation of these depolarizing and hyperpolarizing potentials was observed in addition to monophasic potentials. The presumed EPSPs were 1.0 to 2.8 mV Žmonophasic, 1.3 " 0.2 mV, n s 4. in amplitude and 74 to 93 ms Ž81.0 " 4.3 ms, n s 4. in
Fig. 4. Presumed EPSPs and IPSPs in fetal SC neurons. ŽA. This neuron revealed both presumed EPSPs and IPSPs Žarrows. at V h of y55 mV. Some EPSPs showed temporal summation. ŽB. In this neuron, only IPSPs Žarrows. were observed at V h of y55 mV. One IPSP showed temporal summation. The upper records of ŽA. and ŽB. indicate action potentials evoked by a depolarizing current pulse Ž40 pA..
duration. The amplitude and duration of presumed IPSPs Žmonophasic IPSPs. were 4.0 mV ŽS.E.,0.8 mV, n s 6. and 124.0 ms ŽS.E.,18.0 ms, n s 6., respectively.
4. Discussion This paper describes for the first time in vivo membrane properties of brain neurons of the fetal rat, which is still connected with the dam by the intact umbilical cord. The membrane properties of the fetal SC neurons revealed immaturity in action potentials and resting membrane properties. Although there are no comparable data on the membrane properties of fetal and neonatal SC neurons, Spigelman et al. w33x have reported whole-cell recordings of hippocampal neurons in brain slices obtained from rats of postnatal day ŽPD. 2 to 30. The membrane properties of hippocampal neurons of PD2–5 in this in vitro study resemble those of fetal SC neurons recorded in the present experiments. In the neonatal hippocampal neurons, the mean amplitude of action potentials was approximately 60 mV and the mean duration was about 16 ms ŽTable 1.. They also reported the large R i Ž1100 M V . and long time constant Ž50 ms. for the neonatal hippocampal neurons ŽTable 1.. The resting membrane potential of these immature neurons Žmean s y60 mV. was more depolarized than that of mature neurons Že.g., y72 mV in sliced
Y. Sakata et al.r DeÕelopmental Brain Research 108 (1998) 255–262
hippocampal neurons at PD30. ŽTable 1. w33x. Compared to the neonatal hippocampal neurons, the membrane properties of the fetal SC neurons appear to be more immature. Smaller spike amplitude and longer spike duration in immature brain neurons have been also reported in other brain regions ŽTable 1. w11,15,33,34x. The resting membrane potentials andror the kinetics and density of ion channels underlying spike genesis and repolarization are the likely causes of the smaller spike amplitudes and longer spike durations observed in the present experiments. It is unlikely that the membrane potential is a cause of the small spike amplitudes and longer durations, because they were measured by holding a membrane potential near y60 mV. Thus the kinetics and density of ion channels of fetal SC neurons are thought to be different from those of more matured neurons ŽTable 1. w16,33x. Previous studies have indicated that the smaller spike amplitude of immature brain neurons is due to a low density of functional sodium channels w11,33x. It is likely that the same mechanism underlies the smaller spike amplitude of fetal SC neurons. The lower density of sodium channels may also account for the higher threshold for spike generation observed in fetal SC neurons Žbetween y50 and y35 mV.. In the present experiments, however, it remains to be determined if either the total number of sodium channels or the number of functional sodium channels on fetal SC neurons is smaller as compared to more matured neurons. Since most fetal SC neurons recorded did not reveal afterhyperpolarization, potassium channels involved in the genesis of afterhyperpolarization are considered to be not fully developed in these immature neurons. The lack of afterhyperpolarization may contribute partly to the longer spike duration of fetal SC neurons. In addition, the immaturity of sodium–potassium pumps may also be attributable to the spike configuration of the immature neurons w6x. The slow depolarizations of smaller amplitudes were often observed in fetal SC neurons. These depolarizations with a slow rising and falling phase did not occur at the resting membrane potentials but were evoked by intracellular injection of a long-duration depolarizing pulse. Furthermore, this depolarization occurred with a long delay, when the multiple spikes were induced by a long depolarizing pulse. These findings indicate that the spikelike depolarization is evoked under the state of a long-lasting depolarization. Thus it is likely that the observed depolarizations result from a partial inactivation of sodium channels andror are associated with calcium influx. Immature brain neurons are known to reveal more linear I–V relations than adult neurons, although rectification was often present in both the hyperpolarizing and depolarizing range w19x. Similar features of I–V relations were also observed in fetal SC neurons. The majority Ž7r11. of the fetal SC neurons examined exhibited no anomalous rectification. This finding is consistent with the
261
in vitro studies that reported the absence of anomalous rectifications in hippocampal w31x and neostriatal neurons w20x at early developmental stages. Furthermore, in a developmental study on rat neostriatal neurons using in vivo preparations, Tepper and Trend w34x have demonstrated that in rats younger than PD11, only 13% of neostriatal neurons examined displayed anomalous rectifications and the proportion of neostriatal neurons showing anomalous rectifications increased with development. On the other hand, a small fraction Ž20%. of neostriatal neurons from the first and second postnatal weeks exhibits the outward rectification w34x. Itaya et al. w12x and Molotchnikoff and Itaya w21x reported that the spontaneous activity of SC neurons in rats first appeared at PD5, while electrical stimulation of the optic nerve could not produce evoked response in the SC until PD10 but increased the spontaneous discharge of neonatal SC neurons. Based on these findings, they suggested that despite the presence of spontaneous activity of retinal ganglion cells w18x, classical synaptic transmission did not occur in the rat SC during prenatal periods. In contrast, Reece and Lim reported that in in vitro preparations, optic nerve stimulation could elicit evoked responses in the SC as early as E18 w28x. These responses were blocked by a calcium blocker and a glutamatergic antagonist thus suggesting their postsynaptic origin. Although the discrepancy between the in vivo and the in vitro study remains to be explained, the present experiments, in which presumed postsynaptic potentials were recorded in fetal SC neurons, may support the presence of synaptic transmission in the SC before birth. Furthermore, the higher threshold for spike generation of fetal SC neurons observed in the present experiments may explain the failure of Itaya et al. to obtain the SC responses of the immature brain to optic nerve stimulation. Using the patch clamp recording, Grantyn et al. w8x have found that 44 h after the start of dissociated cell cultures derived from the SC of E21 rats, spontaneous bicuculline-sensitive Cly currents could be recorded. The frequency of the currents could be increased by superfusing the cultured cells with kainate and reduced by addition of blockers of voltage-dependent Ca2q channels. This study suggests that spontaneous IPSPs mediated by GABA occur in the SC at early developmental stages, probably before birth. 5. Unlinked References w2x Acknowledgements The authors are grateful to Dr. H. Onimaru and Dr. T. Shosaku for kind advices on in vivo whole-cell recording technique and for reviewing the manuscript. This work was partly supported by a Grant-in-Aid for Science Re-
262
Y. Sakata et al.r DeÕelopmental Brain Research 108 (1998) 255–262
search ŽNo. 06454475. from the Ministry of Education, Science and Culture of Japan.
References w1x M.G. Blanton, J.J. Lo Turco, A.R. Kriegstein, Whole cell recording from neurons in slices of reptilian and mammalian cerebral cortex, J. Neurosci. Methods 30 Ž1989. 203–210. w2x O.S. Cardoso, K.P. Hoffman, The cortico-tectal projection of the rat in vitro: development, anatomy and physiological characteristics, Eur. J. Neurosci. 7 Ž1995. 599–612. w3x J.B. Dean, E.A. Gallman, D.E. Millhorn, Whole-cell recordings in the dorsal vagal complex ŽDVC. in thick slices prepared from adult rat, in: D.T. Frazier, M.S. Dekin, W.R. Revelette, D.F. Speck ŽEds.., International Conference on Modulation of Respiratory Pattern: Peripheral and Central Mechanisms, Lexington, KY, USA, 1990, pp. 57. w4x F.A. Edwards, A. Konnerth, B. Sakmann, T. Takahashi, A thin slice preparation for patch clamp recordings from neurons of the mammalian central nervous system, Pflug. ¨ Arch. Eur. J. Physiol. 414 Ž1989. 600–612. w5x R.E. Foster, B.W. Conners, S.G. Waxman, Rat optic nerve: electrophysiological, pharmacological and anatomical studies during development, Dev. Brain Res. 3 Ž1982. 371–386. w6x A. Fukuda, D.A. Prince, Postnatal development of electrogenic sodium pump activity in rat hippocampal pyramidal neurons, Dev. Brain Res. 65 Ž1992. 101–114. w7x L. Galli, L. Maffei, Spontaneous impulse activity of rat retinal ganglion cells in prenatal life, Science 245 Ž1988. 90–91. w8x R. Grantyn, K. Krazewski, I. Melnick, H. Taschenberger, S.S. Warton, In vitro development of vertebrate central synapses, Perspect. Dev. Neurol. 2 Ž1995. 387–397. w9x R. Horn, A. Marty, Muscarinic activation of ionic currents measured by a new whole cell recording method, J. Gen. Physiol. 92 Ž1988. 145–159. w10x D.H. Hubel, T.N. Wiesel, Binocular interaction in striate cortex of kittens reared with artificial squint, J. Neurophysiol. 28 Ž1965. 1041–1059. w11x J.R. Huguenard, O.P. Hamill, D.A. Prince, Developmental changes in Naq conductances in rat neocortical neurons: appearance of a slowly inactivating component, J. Neurophysiol. 59 Ž1988. 778–779. w12x S.K. Itaya, S. Fortin, S. Molotchnikoff, Evolution of spontaneous activity in the developing rat superior colliculus, Can. J. Physiol. Pharmacol. 73 Ž1995. 1372–1377. w13x E.R. Kandel, J.H. Schwartz, Directly gated transmission at the nerve–muscle synapse, in: E.R. Kandel, J.H. Schwartz, T.M. Jessell ŽEds.., Principles of Neural Science, Elsevier, Tokyo, 1991, pp. 153–172. w14x F. Kimura, S. Nakamura, Locus coeruleus neurons in the neonatal rats: electrical activity and responses to sensory stimulation, Dev. Brain Res. 23 Ž1985. 301–305. w15x A.R. Kriegstein, T. Suppes, D.A. Prince, Cellular and synaptic physiology and epileptogenesis of developing rat neocortical neurons in vitro, Dev. Brain Res. 34 Ž1986. 161–171.
w16x J. Lopez-Barned, R. Llinas, Electrophysiology of mammalian tectal neurons in vitro: I. Transient ionic conductance, J. Neurophysiol. 60 Ž1988. 853–868. w17x R.D. Lund, A.H. Bunt, Prenatal development of central optic pathways in albino rats, J. Comp. Neurol. 167 Ž1976. 247–264. w18x L. Maffei, L. Galli-Rosta, Correlation in the discharges of neighboring rat retinal ganglion cells during prenatal life, Proc. Natl. Acad. Sci. U.S.A. 87 Ž1990. 2861–2864. w19x D.A. McCormick, D.A. Prince, Postnatal development of electrophysiological properties of rat cerebral cortical pyramidal neurons, J. Physiol. 393 Ž1987. 743–762. w20x U. Misgeld, H.U. Dodt, M. Frotscher, Late development of intrinsic excitation in the rat neostriatum: an in vitro study, Dev. Brain Res. 27 Ž1986. 59–67. w21x S. Molotchnikoff, S.K. Itaya, Functional development of the neonatal retino-tectal pathway, Dev. Brain Res. 72 Ž1993. 300–304. w22x R.D. Mooney, M.Y. Shi, R.W. Rhoades, Modulation of retino-tectal transmission by presynaptic 5-HT1B receptors in the superior colliculus of the adult hamster, J. Neurophysiol. 72 Ž1994. 3–13. w23x S. Nakamura, T. Sakaguchi, Development and plasticity of the locus coeruleus: a review of recent physiological and pharmacological experimentation, Prog. Neurobiol. 34 Ž1990. 505–526. w24x S. Nakamura, S. Arakawa, S. Nishiike, A long-lasting change in the excitability of fetal brain neurons following activation of dam’s hypothalamus in rats, Am. J. Physiol. 269 Ž1995. R236–R244. w25x H. Onimaru, I. Homma, Whole cell recordings from respiratory neurons in the medulla of brainstem–spinal cord preparations isolated from newborn rat, Pflug. ¨ Arch. Eur. J. Physiol. 420 Ž1992. 399–406. w26x H. Onimaru, A. Arata, I. Homma, Intrinsic burst generation of preinspiratory neurons in the medulla of brainstem–spinal cord preparations isolated from newborn rats, Exp. Brain Res. 106 Ž1995. 57–68. w27x D. Purves ŽEd.., Body and Brain, Harvard Univ. Press, Cambridge, MA, 1988. w28x L.J. Reece, C.H. Lim, Optic nerve stimulation elicits evoked responses in superior colliculus of embryonic rats in vitro, Neurosci. Abstr. 21 Ž1995. 1503. w29x T. Sakaguchi, S. Nakamura, Some in vivo electrophysiological properties of locus coeruleus neurons in fetal rats, Exp. Brain Res. 68 Ž1987. 122–130. w30x Y. Sakata, T. Fujioka, Y. Fujii, I. Akimura, S. Nakamura, In vivo whole-cell recordings of superior collicular neurons in fetal rats, Neurosci. Res. 20 Ž1996. S124, Suppl. w31x M. Segal, Developmental changes in serotonin actions in rat hippocampus, Dev. Brain Res. 52 Ž1990. 247–252. w32x D.K. Simon, D.D.M. O’Leary, Responses of retinal axons in vivo and in vitro to position-encoding molecules in the embryonic superior colliculus, Neuron 9 Ž1992. 977–989. w33x I. Spigelman, L. Zhang, P.L. Carlen, Patch-clamp study of postnatal development of CA1 neurons in rat hippocampal slices: membrane excitability and Kq currents, J. Neurophysiol. 68 Ž1992. 55–69. w34x J.M. Tepper, F. Trend, In vivo studies of the postnatal development of rat neostriatal neurons, Prog. Brain Res. 99 Ž1992. 35–50. w35x T.N. Wiesel, D.H. Hubel, Effects of visual deprivation on morphology and physiology of cells in the cat’s lateral geniculate body, J. Neurophysiol. 26 Ž1963. 978–993.