Information processing by nonspiking interneurons: passive
and active properties of dendritic membrane determine
synaptic integration
Masakazu Takahata *, Akira Takashima, Ryou Hikosaka
Di6ision of Biological Sciences,Graduate School of Science,Hokkaido Uni6ersity,Sapporo060-0810,Japan
Abstract
Nonspiking interneurons control activities of postsynaptic cells without generating action potentials in the central nervous system of many invertebrates. Physiological characteristics of their dendritic membrane have been analyzed in previous studies using single electrode current- and voltage-clamp techniques. We constructed a single compart-ment model of an identified nonspiking interneuron of crayfish. Expericompart-mental results allowed us to simulate how the passive and active properties of the dendritic membrane influence the integrative processing of synaptic inputs. The results showed that not only the peak amplitude but also the time course of synaptic potentials were dependent on the membrane potential level at which the synaptic activity was evoked. When the synaptic input came sequentially, each individual input was still discernible at depolarized levels at which the membrane time constant was short due to depolarization-dependent membrane conductances. In contrast, synaptic potentials merged with each other to develop a sustained potential at hyperpolarized levels where the membrane behaved passively. Thus, synaptic integration in a single nonspiking interneuron depends on the value of membrane potential at which it occurs. This probably reflects the temporal resolution required for specific types of information processing. © 2000 Elsevier Science Ireland Ltd. All rights reserved.
Keywords:Nonspiking interneurons; Crayfish; Synaptic integration; Depolarization-dependent membrane conductance
www.elsevier.com/locate/biosystems
1. Introduction
Nonspiking interneurons (NSIs) control the ac-tivity of postsynaptic cells by membrane potential changes associated with synaptic activities, without generating action potentials themselves (Pearson and Fourtner, 1975; Burrows and Siegler, 1976).
Anatomically, they are characterized by an exten-sive dendritic arborization and lack of axonal structure. These physiological and morphological properties of nonspiking interneurons suggest that they function as local information processors (Wilson and Phillips, 1983; Siegler, 1985). Al-though they do not transmit signals to distant parts in the central nervous system, they process informa-tion and control the activity of nearby cells more precisely than spiking cells because the output of NSIs is graded and continuous (Pearson, 1976). * Corresponding author. Tel.:+81-11-7062749; fax:+
81-11-7064923.
E-mail address:[email protected] (M. Takahata).
Recent electrophysiological studies have re-vealed that the dendritic membrane of NSIs pos-sesses depolarization- and hyperpolarization-dependent membrane conductances (Laurent, 1990, 1991; Takahata et al., 1995). The physiolog-ical characteristics of these active membrane con-ductances appear to differ from cell to cell. Thus, NSIs are not simply passive elements in the infor-mation processing conducted by the central ner-vous system. However, it remains unknown how the passive and active properties of the dendritic membrane of NSIs would influence the integrative process of synaptic inputs. The LDS interneuron is an identified NSI in the terminal abdominal ganglion of crayfish (Reichert et al., 1982), in-volved in the local processing of mechanosensory information from the tailfan (Reichert et al., 1983; Krenz and Reichert, 1985). This study investi-gated its response to sequential synaptic inputs at different membrane potential levels using a single-compartmental cell model based on our experi-mental results (Takashima and Takahata, 2000).
2. Materials and methods
2.1. Experimental
Adult crayfish, Procambarus clarkii Girard, of both sexes ranging 7 – 10 cm in body length were used. The abdominal nerve cord including the terminal (sixth) abdominal ganglion was isolated from the rest of the body and pinned to a silicone elastomer-lined chamber with its dorsal side up. The chamber was filled with crayfish saline. The single electrode current- and voltage-clamp tech-niques were applied to nerve cells in the terminal abdominal ganglion under this in vitro condition. Details of electrophysiological experiments are de-scribed elsewhere (Takahashi et al., 1995; Takashima and Takahata, 2000).
2.2. Modeling and simulation
The LDS interneuron extends numerous den-dritic branches on both sides of the terminal abdominal ganglion (Reichert et al., 1982; Fig. 1A). They are connected with a thick transverse
segment (15 – 20 mm in diameter, 150 – 250 mm in length) located over the midline. All electrophysi-ological experiments were carried out on this seg-ment, impaled with a single glass microelectrode. The LDS interneuron was therefore represented by a single compartment that corresponded to the segment. The whole compartment was regarded as internally isopotential. The total membrane area was assumed to be 200 000 mm2according to the results of three-dimensional morphometry using a confocal laser scanning microscope (Hikosaka et al., 1996). The membrane capacitance per unit area was assumed to be 1 mF/cm2 (Rall et al., 1992; Koch, 1999).
The membrane expresses three kinds of depo-larization-dependent conductances, i.e. a sus-tained (gs) and two types of transient (gt1, gt2) conductances, as well as a leak conductance (gleak) which is voltage-independent (Takashima and Takahata, 2000). Experimental analyses have
gested that all of the three kinds of voltage-depen-dent currents are carried by potassium ions. The membrane potential (V) as a function of time was obtained by numerically solving the first-order ordinary differential equation
dV/dt= −(gs(V−EK)+gt1(V−EK)
+gt2(V−EK)+gleak(V−Eleak)
+gsyn(V−Esyn)−Iinj)/Cm
in which EKand Eleak are the equilibrium poten-tials for outward and leak current respectively,Iinj the current injected intracellularly (Fig. 1B). Cm represents the total membrane capacitance calcu-lated from the membrane area.Ekwas−70.0 mV according to the experimental data. Eleak was as-sumed to be the same as the resting potential.gsyn is the synaptic conductance assumed to follow an alpha function with the peak time of 1.0 ms. Esyn is the reversal potential for the synaptic activity. It was assumed to be 0 mV in the case of excitatory input and −80 mV in the case of inhibitory input. The maximal conductance was adjusted in every experiment so that a potential change hav-ing comparable magnitude with the actual synap-tic potential could be reproduced. Each depolarization-dependent conductance was de-scribed by the Hodgkin and Huxley (1952) type equations (Nelson and Rinzel, 1997). Calculation was performed on a Pentium-class PC with the use of a fourth-order Runge – Kutta numerical integrator (time step=0.001 ms).
3. Results
3.1. Physiological characteristics of the LDS interneuron
In the absence of specific sensory stimulus, the LDS interneuron showed spontaneous synaptic activity (Fig. 2A). This activity consisted of synaptic potentials in either depolarizing or hy-perpolarizing direction from the resting potential. It is noted here that not only the peak amplitude but also the time course are different between depolarizing and hyperpolarizing potentials: de-polarizing synaptic potentials are larger and faster than hyperpolarizing ones. Intracellular recording of voltage responses of the interneuron to
con-Fig. 2. Physiological characteristics of the LDS interneuron. (A) Spontaneous synaptic activity of the LDS interneuron recorded on different time scales. (B) Voltage response (upper traces) of an LDS interneuron shown in Fig. 1A to constant current injection (lower traces; 91 – 5 nA). (C) Membrane currents (lower traces) evoked by membrane potential change (upper traces) from the resting potential. Recordings were all made from the same cell shown in Fig. 1A.
3.2. Simulation of membrane responses to current injection
Fig. 3A illustrates voltage responses of the model to current injection. Comparison of the calculated result with the recorded response (Fig. 2B) suggested that the model well reproduced response characteristics of the real LDS interneu-ron. An exception was the inward rectification upon strong hyperpolarization that was not incor-porated into the current model. We observed
clear-cut outward rectification upon depolariza-tion from the resting potential. A characteristic of this model response was an initial overshoot in the voltage response to large depolarizing current as indicated by an arrowhead in Fig. 3A. The real cell also shows this overshoot (Fig. 2C of Taka-hashi et al., 1995). However, it was absent in some cases (Fig. 2B). Because the same transverse seg-ment was always impaled by the electrode, this variability in the overshoot was probably due to a difference in the balance among sustained and transient conductances that was either physiologi-cal or artifactual due to microelectrode impale-ment. In any case, the observed overshoot in the model response appears to be consistent with the physiological response of the real cell.
3.3. Simulation of synaptic acti6ity
The result of calculation shown in Fig. 3A also illustrates how the shape of a single, depolarizing synaptic potential is influenced by the membrane potential level at which it is evoked. The peak amplitude was increased by hyperpolarization and decreased by depolarization. The time course of the synaptic response was also affected by the membrane potential: the half-decay time of the response was increased by hyperpolarization and decreased by depolarization. We normalized the peak amplitude and half decay time obtained in the present calculation to the values at the resting potential and plotted against membrane potential (Fig. 3B) together with those obtained in physio-logical experiment (Takahata et al., 1995). The general tendency was common to calculated and observed synaptic responses. Decrease in the peak amplitude and the half decay time by membrane depolarization was more remarkable in calculated responses than in observed ones. This discrepancy was partly accounted for by difficulty in measur-ing these parameters at depolarized levels due to the highly reduced size of synaptic activity. A characteristic feature in the observed response was that both the peak amplitude and the half decay time began to decrease when the hyperpolariza-tion became larger. This decrease was due to the inward rectification of the interneuron membrane activated by a large hyperpolarization (Fig. 2B). Fig. 3. Synaptic activity at varying membrane potential levels.
Because such inward rectification property was not incorporated in the present model, the peak amplitude of the synaptic potential continued to increase as the electromotive force increased whereas its time course became constant at hyper-polarized levels (Fig. 3B).
When the synaptic input occurred sequentially, it made temporal summation with the following one to integrate successive signals. This integra-tion was significantly affected by the membrane potential level at which the synaptic activity was evoked. A train of depolarizing input (100 Hz) was simulated at various membrane potential lev-els in Fig. 4A. The maximal synaptic conductance was adjusted so that the peak amplitude of the first potential was almost comparable in each case. The results showed that individual synaptic input was better preserved as the membrane was more depolarized. At resting and hyperpolarized levels, individual synaptic potential merged with each other to develop a rather sustained potential so that information on individual input was al-most lost. Similar tendency was observed in the case of hyperpolarizing input (Fig. 4B). Discrimi-nation of individual potential at depolarized levels was less remarkable than in the case of depolariz-ing input. But it was clear that information on individual input is better preserved at depolarized membrane potential levels. It is thus suggested that the temporal resolution in synaptic activity critically depends on the membrane potential level due to depolarization-dependent conductances.
Temporal summation of synaptic potentials at the resting potential is compared between depo-larizing and hyperpodepo-larizing inputs in Fig. 4C. The maximal synaptic conductance for each input was adjusted so that the peak amplitude of calcu-lated synaptic potentials was comparable with that of the recorded response in physiological experiments (Takahata et al., 1995). The same input interval value was used for both activities. Our results showed that information on individual input was better preserved in the depolarizing activity than in the hyperpolarizing one. The tem-poral resolution in the synaptic integration there-fore depends on the polarity of synaptic inputs even at the resting potential level in the LDS interneuron.
Fig. 4. Temporal synaptic summation at varying membrane potential levels. (A) A train of depolarizing synaptic inputs (100 Hz) at different membrane potential levels. The maximal synaptic conductance was 0.7 nS for 5 nA injection, 0.4 nS for 2 nA, 0.3 nS for 1 nA, 0.2 nS for 0 nA, 0.15 nS for −1 nA, 0.1 nS for −3 nA and −5 nA injection. (B) A train of hyperpolarizing inputs at the same frequency. The maximal synaptic conductance was 0.7 nS for 0 nA, 2 nA, 5 nA, 2 nS for−1 nA injection. (C) A train of depolarizing and hyperpo-larizing inputs at the resting potential. Stimulus intensity for the depolarizing response was greater than in (A) so that the rapid voltage decay was observed clearly.
4. Discussion
of information processing by NSIs to quantita-tively investigate how the synaptic input is inte-grated and transformed into output to postsynaptic cells. In this study, we examined how the physiological properties of the LDS interneu-ron (Takashima and Takahata, 2000) would influ-ence the synaptic integration in the cell using a single compartment model.
Voltage responses of the model to constant current injection (Fig. 3A) were very similar to those observed in physiological experiments (Fig. 2B) showing outward rectification upon depolar-ization from the resting potential with initial over-shooting (Takahashi et al., 1995). Calculation of synaptic potentials at different membrane poten-tial levels has revealed that not only their peak amplitude but also the time course were depen-dent on the membrane potential level at which the synaptic activity was evoked. The calculated change in the shape of synaptic potentials coin-cided with the experimental data (Fig. 3B). The exception was observed at strongly hyperpolarized levels where membrane showed inward rectifica-tion that was not incorporated into the present model.
When either depolarizing (Fig. 4A) or hyperpo-larizing (Fig. 4B) synaptic inputs occurred sequen-tially, the individual inputs were still discernible at depolarized levels at which the membrane time constant was short due to depolarization-depen-dent membrane conductances. In contrast, synap-tic potentials merged with each other to develop a sustained potential at hyperpolarized levels where the membrane behaved passively. At the resting potential level, sequential depolarizing synaptic input tends to preserve the shape of individual potential whereas hyperpolarizing input makes temporal summation with the following ones to develop a sustained potential change (Fig. 4C). The synaptic integration in the LDS interneuron thus depends on the membrane potential level at which the synaptic activity is evoked as well as the polarity of synaptic inputs when they are evoked at the resting potential level.
The LDS interneuron establishes monosynaptic excitatory connection with mechanosensory affer-ents (Reichert et al., 1982) showing a large depo-larizing potential (10 – 30 mV) upon natural or
electrical stimulation (Reichert et al., 1983). The interneuron responds in a 1:1 manner to water current stimulation of 2 Hz and inhibit postsynap-tic cells effectively (Reichert et al., 1983). The short membrane time constant would be advanta-geous for the LDS interneuron as a mediator of lateral inhibition in the tailfan mechanosensory system since it enhances the time resolution of the system that handles water current stimuli coming in succession one after another. However, this possibility has to be experimentally examined by applying blockers for active conductances and observing their effects on the mechanosensory system function. In contrast, the long membrane time constant has been suggested in premotor NSIs to be advantageous for transforming a train of discrete synaptic input into a sustained, contin-uous output for controlling the general excitabil-ity of postsynaptic motoneurons (Murayama and Takahata, 1998). It remains open to future study, however, what the functional significance of sus-tained hyperpolarizing potential is for the LDS interneuron since it is only known to inhibit post-synaptic interneurons by depolarizing post-synaptic ac-tivity although it receives hyperpolarizing inputs (Fig. 2A).
Acknowledgements
This work was funded in part by grants from Japan Society for the Promotion of Sciences (AT, RH) and Grants-in-Aid (09440274, 11168202) from the Ministry of Education, Science, Sports and Culture of Japan (MT).
References
Burrows, M., 1979. Synaptic potentials effect the release of transmitter from locust non-spiking interneurons. Science 204, 81 – 83.
Burrows, M., Siegler, M.V.S., 1976. Transmission without spikes between locust interneurones and motoneurones. Nature 262, 222 – 224.
Hikosaka, R., Takahata, M., 1998. Quantitative analyses of anatomical and electrotonic structures of crayfish nonspik-ing interneurons by three-dimensional morphometry. J. Comp. Neurol. 392, 373 – 389.
Hikosaka, R., Takahashi, M., Takahata, M., 1996. Variability and invariability in the structure of an identified nonspik-ing interneuron of crayfish as revealed by three-dimen-sional morphometry. Zool. Sci. 13, 69 – 78.
Hodgkin, A.L., Huxley, A.F., 1952. A quantitative description of membrane current and its application to conduction and excitation in nerve. J. Physiol. 117, 500 – 544.
Koch, C., 1999. Biophysics of Computation. Oxford Univer-sity Press, New York.
Krenz, W.D., Reichert, H., 1985. Lateralized inhibitory input to an identified nonspiking local interneuron in the crayfish mechanosensory system. J. Comp. Physiol. 157, 499 – 507. Laurent, G., 1990. Voltage-dependent nonlinearities in the
membrane of locust nonspiking interneurons, and their significance for synaptic integration. J. Neurosci. 10, 2268 – 2280.
Laurent, G., 1991. Evidence for voltage-activated outward currents in the neuropilar membrane of locust nonspiking local interneurons. J. Neurosci. 11, 1713 – 1726.
Mizunami, M., Yokohari, F., Takahata, M., 1999. Explo-ration into the adaptive design of the arthropod ‘micro-brain’. Zool. Sci. 16, 703 – 709.
Murayama, M., Takahata, M., 1998. Neuronal mechanisms underlying the facilitatory control of uropod steering be-haviour during treadmill walking in crayfish. I. Antagonis-tically regulated background excitability of uropod motoneurones. J. Exp. Biol. 201, 1283 – 1294.
Nelson, M., Rinzel, J., 1997. The Hodgkin – Huxley model. In: Bower, J.M., Beeman, D. (Eds.), The Book of GENESIS, 2nd edn. Springer – Verlag, New York, pp. 29 – 49. Pearson, K.G., 1976. Nerve cells without action potentials. In:
Fentress, J.C. (Ed.), Simpler Networks and Behavior. Sin-auer, Sunderland, pp. 99 – 110.
Pearson, K.G., Fourtner, C.R., 1975. Nonspiking interneurons in walking system of the cockroach. J. Neurophysiol. 38, 33 – 52.
Rall, W., Burke, R.E., Holmes, W.R., Jack, J.J.B., Redman, S.J., Segev, I., 1992. Matching dendritic neuron models to experimental data. Physiol. Rev. (Suppl.) 72, 159 – 186. Reichert, H., Plummer, M.R., Hagiwara, G., Roth, R.L.,
Wine, J.J., 1982. Local interneurons in the terminal ab-dominal ganglion of the crayfish. J. Comp. Physiol. 149, 145 – 162.
Reichert, H., Plummer, M.R., Wine, J.J., 1983. Identified nonspiking local interneurons mediate nonrecurrent, lat-eral inhibition of crayfish mechanosensory interneurons. J. Comp. Physiol. 151, 261 – 276.
Siegler, M.V.S., 1985. Nonspiking interneurons and motor control in insects. Adv. Insect Physiol. 18, 249 – 304. Takahashi, M., Takahata, M., 1995. Dendritic properties of
uropod motoneurons and premotor nonspiking interneu-rons in the crayfishProcambarus clarkiiGirard. J. Comp. Physiol. 176, 503 – 512.
Takahashi, M., Takashima, A., Takahata, M., 1995. Regional characteristics of the membrane response of an identified crayfish nonspiking interneuron to intracellularly injected current. J. Neurophysiol. 74, 2242 – 2250.
Takahata, M., Niwa, M., Nakamura, H., 1995. Physiological characteristics of the synaptic response of an identified sensory nonspiking interneuron in the crayfish Procam
-barus clarkiiGirard. J. Comp. Physiol. 176, 737 – 746. Takashima, A., Takahata, M., 2000. Electrophysiological and
theoretical analysis of depolarization-dependent outward currents in the dendritic membrane of an identified non-spiking interneuron in crayfish. J. Comput. Neurosci. 9, 187 – 205.
Wilson, J.A., Phillips, C.E., 1983. Pre-motor non-spiking in-terneurons. Prog. Neurobiol. 20, 89 – 107.
