- Email: [email protected]
Single cell isolation: A way to examine network interactions
Brain Research, 179 (1979) 219-230 © Elsevier/North-Holland Biomedical Press
219
S I N G L E CELL ISOLATION: A WAY TO E X A M I N E N E T W O R K INTERACTIONS
J. A. LONDON and MICHAEL MERICKEL Department of Physiology and Biophysics, 524 Burrill Hall, University of Illinois, Urbana, Ill. 61801 (u.s.A.)
(Accepted April 12th, 1979) Key words:
helisoma - - network - - cell isolation
SUMMARY A procedure for isolating identified, small neurons from snail ganglia is described. The technique allows a particular neuron, previously identified by morphological and electrophysiological characteristics, to be marked and then isolated from the ganglia. This procedure was developed to permit the detailed comparison of the electrical characteristics of a neuron before and after isolation from an intact system. An earlier description has appeared is. The cell somata is marked intracellularly by the iontophoretic injection of Procion navy blue H3RS which visually differentiates the cell from other cells in the ganglion. The ganglion is then treated with a trypsin-haluronidase solution to soften the ganglion sheath, which is then removed. The cells are gently shaken to isolate them from the ganglion and then examined electrophysiologically. A comparison of membrane properties, such as action potential height, duration and rate of rise and decay was made before and after all treatments were applied to assess deleterious effect. An analysis of network properties, such as burst duration, number of spikes per burst and presynaptic activity was also performed after each phase of the procedure. No significant differences were noted after dye injection, enzyme treatment, and where appropriate, after isolation. An increase in input resistance and corresponding decrease in the slope of the steady state current-voltage plot (I-V plot) were observed after isolation of a cell. These were expected results of removing the 'load' (i.e. axon or electrical coupling) from the cell soma. This method may be applied to many other systems to study the effects of network interactions on the properties of a single cell and should therefore facilitate the analysis of neuronal networks as well as single cell properties.
220 INTRODUCTION It is becoming increasingly apparent that individual cell properties are important in determining the generation of activity by a network 23. It is therefore necessary to study a neuron in isolation in order to examine its specific membrane properties separately from those electrical and chemical synaptic interactions with other neurons. This paper describes a single cell isolation procedure applicable to small, identified neurons. Some separation of network properties from intrinsic properties has been accomplished using a variety of techniques - - most notably chemical and physical isolation. The use of EDTA is a commonly used chemical method of isolating cells from synaptic inputL However, the reduction of calcium in the solution as a result of the chelating action of EDTA can be expected to affect Ca-dependent membrane properties 6. Another method of isolation from synaptic input is replacement of Ca in the Ringer's solution with magnesium, which has the same problem. Physical isolation of neurons was first accomplished by Oomura and Maeno t,9. The isolated cells of this study exhibited a decrease in membrane potential and did not show spontaneously firing action potentials upon penetration with a microelectrode, indicating possible damage to the cells. Other isolation procedures have been suggested, but these are confined to studying the largest cells in the ganglion. Usually such techniques do not allow a small cell to be studied in the intact system and in isolation due to the difficulty of following the cell through the isolation procedure. Therefore, averaging of the parameters under investigation from many cells is often utilized instead of studying the same cell or cells before and after isolationl,3,4,9,1~, 17. This investigation described two related techniques: (1) a method for isolating a cell from a ganglion with minimal damage; and (2) a method for marking a particular cell, thus distinguishing it from the rest of the cells in the ganglion, allowing the cell to be identified after isolation. These two techniques permit the properties of the same cell to be studied while in a network of cells and after being removed. The experimental paradigm consisted of comparing various parameters, before and after isolation, which were chosen to reflect both the integrity of individual cells and synaptic interactions. The medial giant cells (cells 5) and the main pair of protractor neurons (cells 19) were the subjects of this study I1. The medial giant cells (75-100/~m in diameter) have little synaptic input, and are only distantly electrically coupled. The protractor neurons (approximately 50/~m in diameter) are members of a group of electrically coupled motoneurons which receive large, rhythmic IPSPs from the cyberchron network, a group of rhythmically bursting cells T M . The cyberchron cells are also an electrically coupled group of cells whose output drives the motoneurons (e.g. protractor motoneurons, such as 19 and retractor motoneurons) involved in the feeding behavior of the snail. Therefore, monitoring the large IPSPs on cells 19 also serves as an indicator of the integrity of the cyberchron network as well as that of cells 19.
221 MATERIALSAND METHODS
Specimen preparation Specimens of the fresh-water snail, Helisoma trivolvis albino, were maintained in a well-aerated aquarium and kept on a 12 h light/dark cycle. Details of animal care, dissection of the buccal mass (BM) and buccal ganglion (BG) and preparation immobilization for recording have been previously described in Kater et al. la,14. Cells were identified according to the system of Kater 14,15. All experiments and dissections were performed with the preparation immersed in physiological saline consisting of: 51.3 mM NaC1, 1.7 mM KCI, 4.1 mM CaCI2, 1.5 mM MgCI2, and 1.8 mM NaHCOa added as a buffer (pH 7.6) (see refs. 11-13).
Electrical recording Glass microelectrodes filled with 2 M potassium citrate (DC resistance = 20-40 Mf~) were used for intracellular recording, and an Ag-AgCI wire served as the indifferent electrode. A high input impedance preamplifier with a bridge circuit for injecting current (Getting Microelectrode Amplifier, Los Altos, Calif., Model 4) was used for intracellular recording and signals were displayed on a Tektronix 5000 series oscilloscope and on a Brush (Model 280) chart recorder. Action potentials were photographed from the oscilloscope using a Tektronix (Model C5) oscilloscope camera. Microelectrodes for dye injection were filled with a 4 % solution of Procion navy blue H3RS (PNB) (Polysciences), a negatively charged dye, in distilled water (DC resistance : 60-100 Mf~). PNB was injected into the cell by passing hyperpolarizing current of 1-10 nA through the electrode until a small area of blue appeared in the cell. The duration of current application varied with the electrode characteristics and the amount of applied current, but never exceeded 5 min. This blue area remained visible in the cell for at least 10 h. Cells marked in this manner were readily distinguishable from other cells in the ganglion even when all morphological characteristics (such as position and relative size of the cell in the ganglion) and characteristic electrical activity were removed by the cell's isolation from the ganglion. Cell membrane parameters were measured before and after each experimental manipulation to examine effects on the cell and included: (1) membrane potential; (2) voltage response to a step input current from which time constants, input resistance and steady-state I-V plots were determined; and (3) action potential characteristics, such as height, duration, and rates of rise and fall. Network interactions were assessed before and after each experimental manipulation by measuring: (1) feeding cycle time (defined as the number of bursts occurring in a given time recorded in protractor motoneuron 19); and (2) the shape of the IPSPs recorded in protractor motoneuron 19.
The isolation procedure The cell was marked with PNB after all initial measurements and experiments were performed. The BG and BM were then softened with the enzyme digestive
222 treatment which consisted of the following steps: (1) a 5 min soak in 0.12 % solution, by weight, of trypsin-hyaluronidase (Calbiochem) dissolved in saline, at 22--25 C , softening the sheath of the ganglion to facilitate the sheath's removal; and (2) two 10 min washes of saline at 5 °C. The BM and BG were either repinned in the dish used for electrical recording and data was again taken, or the isolation procedure continued with the BG being separated from the BM and the former placed in the cell isolation apparatus. The cell isolation apparatus consisted of a small petri dish whose bottom was covered with a layer of Sylgard, and then a layer of 0.5 % Agar (Difco, dissolved in saline). The cells clung to the surface of the Agar when they were isolated, immobilizing them and permitting penetration with a microelectrode. The petri dish was placed in a larger dish filled with ice which increased the number of viable cells and their lifespan. After pinning the BG, its sheath was cut away using electrolytically sharpened Tungsten needles. Gentle shaking of the ganglion isolated the cells from each other and from the neuropile. By varying the vigor of shaking, the amount of remaining axon could he controlled to some extent. The indifferent electrode was then placed in the dish containing the isolated cells, and standard electrical recordings (see Electrical recording) were taken. A cell which had been stained with PNB was easily visually differentiated from surrounding cells permitting comparative studies to be conducted. RESULTS
Effect of Procion navy blue A comparison of the activity of a cell before and after injection of PNB (cell 19; Fig. 1) shows very little change in action potential characteristics, and a small change in presynaptic activity and cycle time. There was an average change of 6.0 ~ for 10 preparations. Comparisons of the cell's input resistance and time constants (Fig. 1) before and after dye injection also shows little change. The membrane potential of the cell before intracellular marking was --60 mV, compared to --58 mV after, indicating B-PNB
,
,
i _
_
Im2v
J
0.5sot Fig. 1. Activity of cell 19 before and after PNB injection. B-PNB: before dye injection. AoPNB: after dye injection. Top traces show activity of the cell as a result of input from the network. Bottom traces show cell response to a small, hyperpolarizing step current injected.
223
CURRENT (nA)
0.8 0.6
•
0.4
•
0.2 "A I --30
I -20
I
I -10
10
-0.2
I 20
I
30
VOLTAGE
(mY)
-0.4 &Q -0.6
-0.8 Fig. 2. Steady-state 1-V plot of a cell (cell 19) before and after PNB injection. In all I-V plots shown, the plot was constructed by measuring the voltage response of the cell to different step current injections. A : before injection; • : after injection. a very small effect o f the dye. In 10 other cells examined, the m e m b r a n e potential never changed by more than ± 2 mV after dye injection (in 4 cases Em became more negative, in 5 cases more positive, and in 1 case remained unchanged). It was also noted that the cell could still fire out o f anode break after being marked with the dye, providing additional evidence that the cell was not affected by the dye injection. The steady-state I - V plot (Fig. 2) o f a cell (cell 19) before and after dye injection indicates that the steady-state m e m b r a n e properties were not affected by this
TH
~
~.'
Before
After
2O
ms
Fig. 3. Examination of the action potential of a cell before and after a given treatment. TH: before and after trypsin-hyaluronidase treatment (cell 19). PNB: before and after PNB injection. Top trace of each inset shows the voltage response of the cell. Bottom trace of each inset shows the current trace. The treatments were given separately to two diffrent cells.
224
I-
i
,J4omV 2sec
L___I
I nA 0.5 sec
Fig. 4. Cell 19 activity before and after trypsin-hyaluronidase treatment. B-TH: before application of the enzymes. A-TH: after application of the enzymes and washes. Top traces show activity of the cell as a result of input from the rest of the network. Bottom traces show cell response to a small, hyperpolarizing step current injected.
treatment. The passive membrane region appears to be linear before and after treatment. A closer look at an action potential of a marked cell (cell 19, Fig. 3) indicates that action potential duration, threshold and height were not significantly altered. The above tests demonstrate that the passive membrane properties and those active membrane properties contributing to action potential generation do not appear to be affected by PNB injection.
CURRENT (r~A) L0'
t
0.8" 0.6'
•
0.4" 0.2-
-30
-16 •'
-2o
-0.2•
•
*
O
1'o
2b
3o
VOLTAGE (mY)
-0.4"
-0.6.
Fig. 5. Steady-state I-V plot of a ~11 (cell 5) before and after trypsin-hyaluronidase treatment. ~ , before treatment; O, after treatment. For an explanation of how points were obtained, see Fig. 3 legend.
225
Effect of trypsin-hyaluronidase treatment The same parameters used to analyze the extent of damage in the PNB-injected cells were used to assess the effects of the enzymatic treatment, i.e those of network interaction and of cell membrane response. Action potential characteristics, synaptic activity, and IPSP frequency were not significantly affected by the enzyme treatment (cell 19, Fig. 4). A comparison of passive membrane properties also showed that input resistance and time constant were not noticeably affected. The passive membrane region (where I-V plot is linear) and region of rectification exhibited little change after treatment (cell 5, Fig. 5). Comparison of action potential characteristics (e.g. height, duration, and rates of rise and decay) (Fig. 3) before and after the enzymatic treatment provides further evidence that little alteration has occurred in the cells as a result of this treatment. The resting potential of only one cell out of ten different cells examined was slightly affected (more positive by 3 mV). Delayed and combined treatment effects Possible deleterious effects due to the combination of both PNB injection and enzymatic digestion were investigated. The cell was examined at 2, 4, 6 and 8 h after application of both treatments in order to demonstrate possible long term effects. A comparison of the I-V plots of the cell before and 8 h after administration of the treatments showed insignificant differences in both the rectification and the passive linear regions of the steady-state I-V plots (Fig. 6). The isolated cell After analyzing and marking a cell and enzymatically treating the BG and BM the cell was physically isolated from the rest of the cells in the ganglion. Potential deleterious effects from the isolating procedure were assessed using both morphological observations and electrical measurements. As described in Methods (the isolation procedure), cells were isolated by gently shaking the slit ganglion in fresh saline.
CURRENT • 0.3~A) 0'2! 0.1t . :zG -,'s -lb -;, I[ ae
•A A •&
•
$
-0.1
-0.2 &
-0.3 Fig. 6. Steady-state I - V plot of a cell (cell 5) before both a PNB injection and a T H treatment were administered to the same cell, and 8 h after both treatments were administered. A , before treatment; Q, after treatment. For an explanation of how points were obtained for this plot, see the legend of Fig. 3.
226
Fig. 7. A series of pictures of 4 different ganglia taken at various stages of the isolation procedure. A : two different cells have been injected with PNB. Injected cells are marked with arrows. Note how the dye is contained inside the ceil. Dye stays visible for a minimum of 8 h after injection. B: the ganglion has been separated from the rest of the preparation, has been cut circumferentialty, and the cells have started to emerge. C: the sheath has been lifted away from the main body of cells, and the ganglion has been lightly shaken. Note that several cells are already completely isolated. D: a completely isolated cell. Bits of glia are still clinging to the axon as evidence of the gentleness of the techniuqe. Also note that the membrane integrity has been maintained, and that the nucleus is small and round, more evidence that this cell is healthy. All calibration marks equal to 100/ml.
M o r p h o l o g i c a l e x a m i n a t i o n of an isolated cell (Fig. 7) d e m o n s t r a t e d the gentleness of this technique in r e m o v i n g a cell from the ganglion. The cell retained a length of axon over 100 # m long, a n d on the axon clinging fragments o f glial cells were visible. The nucleus of the cell was r o u n d a n d small, similar to its a p p e a r a n c e in an unisolated cell. Lastly, the m e m b r a n e of the soma appeared to have m a i n t a i n e d its integrity with n o a p p a r e n t or visible 'blebs' ( m e m b r a n e outpocketings) due to m e m b r a n e degeneration. Inspection o f an action potential (cell 19, inset, Fig. 8) of a cell before a n d after its removal from the ganglion shows insignificant effects on height, d u r a t i o n , rates of rise a n d decay. The I - V curves of this cell before a n d after
CUI~nR~NT 1.0. ,:
O.8
B-I
0.6-
',.____
"~
•
0.4-
A-I
0.2-
-~'o
•
•J
-2b -10**•
1'0
2'0
3"0 4"0 VOLTAGE
-~2
•
-0.4 -0.6
Fig. 8. Steady-state I - V plot of a cell (cell 19) before and after isolation. Insets show the action potential of the cell before and after isolation• Top trace of the insets is the voltage response of the ceil. Bottom trace of the insets shows the current injected. A , before isolation; O, after isolation•
• O.E
~....
•
A-I
"-~
--~
~'...
",, ' "" >"I 0.(
", -...'%,
"...-
-0~
* ---l %.."~Q ~.."%,
"A''i B-I
--I.
7
~
l.'v '
V-- 10.2 ~lnll l
i""e'-,~
.~----
-,,,
"z.,, "-*,
-L6 ---I
1'o
'
J
3'0
",.,..
'
sb
TIME (msec)
'
7'0
'
9'0
Fig• 9. Comparison of time constants before and after isolation of cell 19. This graph illustrates that the non-isolated cell response to a small hyperpolarizing current pulse has two components: the fast component with the steep slope is due to the axon; and a slow, gradual sloped component (due to the cell soma)• The isolated cell has only one component - - the slow soma response. The insets show the response of the same cell before and after isolation to the same size current injection• B-I, before isolation; A-I, after isolation. The top trace of each inset shows the membrane voltage response and the bottom trace is the current injected.
228 isolation, however, showed a change in the slope of the passive linear region of the celt. The input resistance increased as expected with isolation of the cell. The two regions of rectification (hyperpolarizing and depolarizing) were rotated due to the general increase in steady-state resistance. A more detailed look at the passive membrane properties was obtained by examining a graph o f L n (Vr-V(t)) vs time (cell 19, Fig. 9). Various linear regions of this graph correspond to different time-constant processes. Two time-constant processes were observed from the intact cell in the ganglion: (1) a slow time-constant (90 msec) corresponding to the soma time constant; and (2) a faster time-constant (33 msec) due to the properties of the axon 7. In the isolated cell, one time-constant was seen. The faster time-constant was abolished as a result of removing most of the axon. DISCUSSION Procion navy blue injection As has been previously described, a cell in a ganglion can be identified by its characteristic electrical activity, its relative morphological location and its size. However, when a cell is isolated these criteria can no longer be applied to most cells, and a direct comparison of a cell's properties before and after isolation cannot be made. A way to overcome this difficulty is to physically mark the cell to be studied. This study utilized the dye Procion navy blue H3RS which we have shown to have insignificant effects on several membrane properties. However, there has been some evidence reported that the Procion dyes affect the membrane properties of the cell, and also the cell's internal structuresS, 14,2°. There are several possible explanations for these discrepancies. In earlier studies the dye was injected for several hours, increasing the amount of dye injected and applied current. Enzymatic treatment Many studies have demonstrated the deleterious effects of digestive enzymes on membrane properties3,6,2L An increase in action potential duration, membrane depolarization and inactivation of specific currents after prolonged exposure to digestive enzymes such as trypsin, hyaluronidase and pronase has been reported 6,~z. In this study, by carefully modulating solution temperature and the application time of the enzymes, the damaging effects of the enzymes were minimized. No significant differences were seen in individual cell parameters and synaptic interactions between neurons. It is important to note that the exposure time of the preparation to the enzymes was considerably shorter than that of previous studies ~,6,22 (5 min compared to several h). In this study the enzymes were used to soften, but not dissolve the sheath, allowing it to be cut easily without mechanically stressing the cells in the ganglion, other studies applied the enzymes to the ganglia for lengths of time causing the sheath to fall apart 16. The isolation This study makes a direct comparison of the membrane and network properties
229 of a cell that is still connected to other cells in the ganglion and the same cell after being physically isolated from other cells. The isolation technique was shown to be very gentle by use of morphological evidence (Fig. 8) and by comparisons of some relevant membrane and network properties, action potential characteristics, and the rectification areas of the I - V plot, which appear to be unaffected by the isolation procedure. Previous methods have employed large cells and have used the average value from several cells to determine the effects of isolating a celp6,17. Several of these studies used cells that exhibited some degree of mechanical damage, produced either by cutting a sheath that was not softened or by micropipetting the cells from the neuropile during the isolation process16, is. In the present series of experiments it was noted that pipetting the cells resulted in some damage to the cell, and decreased the number of viable ceils, Gently shaking the ganglion to free the cells prevents the stress of the cells bumping into the glass and the turbulence when drawn into the pipette. The effects of loading of the network were demonstrated by comparing the voltage response before and after isolation. It was seen that the axon of the cell can cause at least a 3-fold decrease in input resistance. This decrease was also seen in the change of slope of the passive linear region of the I - V plot. This change in input resistance was compared to the expected value based on the measured time constants (soma and axon) of the intact neuron and from a simple equivalent model 7. A 5-fold increase in resistance is predicted from the model, which is compatible with the 3-fold increase observed. The fact that the calculated increase based on the model is slightly greater than the observed increase is probably due to the fact that a small length of axon remained after isolation. This isolation technique allows a direct comparison of membrane and network parameters of a specific cell while the cell is still in the network, and once it is removed, the interactions of a cell with its surroundings can then be directly evaluated.
REFERENCES 1 Alving, B. O., Spontaneous activity in isolated somata of Aplysia pacemaker neurons, J. gen. Physiol., 45 (1968) 29-45. 2 Asada, Y. and Bennett, M. V. L., Experimental alteration of coupling resistance at an electrotonic synapse, J. Cell BioL, 49 (1971) 159-183. 3 Chen, C. F., von Baumgarten, R. and Takeda, R., Pacemaker properties of completely isolated neurones in Aplysia californica, Nature New Biology, 233 (1971) 27-29. 4 Faber, D. S. and Klee, M. R., Membrane characteristics of bursting pacemaker neurones in Aplysia, Nature New Biology, 240 (1972) 29-31. 5 Getting, P. A., Modification of neuron properties by electrotonic synapses. I. Input resistance, time constant and integration, J. Neurophysiol., 37 (1974) 846-857. 6 Heyer, C. B. and Lux, H. D., Properties of a facilitating calcium current in pace-maker neurones of the snail, Helixpomatia, J. Physiol. (Lond.), 262 (1976) 319-348. 7 Jack, J. B., Noble, D. and Tsein, R. W., Electric Current Flow in Excitable Cells, Clarendon Press, Oxford, 1975, pp. 160-172. 8 Jankowska, E. and Lindstrom, S., Morphological identification of physiologically defined neurones in the cat spinal cord, Brain Research, 20 (1970) 323-326. 9 Kado, R. T., Aplysia giant cells: soma-axon voltage clamp current differences, Science, 182 (1973) 843-845.
230 10 Kaneko, C. R. S., Merickel, M. and Kater, S. B., Centrally programmed feeding in Helisoma controlled by electrically coupled network. 1. Premotor neuron identification and characteristics, Brain Research, in press. 11 Kater, S. B., Feeding in Helisoma trivolvis: the morphological and physiological bases of a fixed action pattern, Amer. ZooL, 14 (1974) 1017-1036. 12 Kater, S. B., Heyer, C. and Hegmann, J. P., Neuromuscular transmission in the gastropod mollusc Helisoma trivolvis: identification of motoneurons, Z. vergl. Physiol., 74 (1971) 127-139. 13 Kater, S. B. and Kaneko, C. R. S., An endogeneously bursting neuron in the gastropod mollusc, Helisoma trivolvis: characterization in vivo, J. comp. Physiol., 79 (1972) 1-14. 14 Kater, S. B. and Nicholson, C., Intracellular Staining in Neurobiology, Springer-Verlag, New York, 1973, pp. 211-226. 15 Kater, S. B. and Rowell, C. H. F., Integration of sensory and centrally programmed components in generation of cyclical feeding activity of Helisoma trivolvis, J. Neurophysiol., 37 (1973) 142--155. 16 Kostenko, M. A., Geletyuk, V. I. and Veprintsev, B. N., Completely isolated neurons in the mollusc Lymnaea stagnalis. A new objective for nerve cell biology investigation, Comp. Biochem. Physiol., 49 (1974) 89-100. 17 Kostyuk, P. G., Krishtal, O. A. and Pidoplichko, V. 1[.,Effect of internal fluoride and phosphate on membrane currents during intracellular dialysis of nerve cells, Nature (Lond.), 257 (1975) 691-693. 18 London, J. A. and Merickel, M. B., Analysis of neural network components using single cell isolation, Neurosci. Abstr., II1,383. 19 Oomura, Y. and Maeno, T., Does the neurone somata actually generate action potentials?, Nature (Lond.), 197 (1963) 358 359. 20 Payton, B. W., Histological staining properties of Procion yellow M4RS, J. Cell BioL, 45 (1970) 659-662. 21 Purves, D. and McMahan, U. J., The distribution of synapses on a physiologically identified motor neuron in the central nervous system of the leech, J. CellBiol., 55 (1972) 205-220. 22 Rojas, E. and Armstrong, C. M., Sodium conductance activation without inactivation in pronaseperfused axons, Nature New Biology, 229 (1971) 177-178. 23 Russell, D. F. and Hartline, D. K., Bursting neural networks: a reexamination, Science, 200 (1978) 453456.
219
S I N G L E CELL ISOLATION: A WAY TO E X A M I N E N E T W O R K INTERACTIONS
J. A. LONDON and MICHAEL MERICKEL Department of Physiology and Biophysics, 524 Burrill Hall, University of Illinois, Urbana, Ill. 61801 (u.s.A.)
(Accepted April 12th, 1979) Key words:
helisoma - - network - - cell isolation
SUMMARY A procedure for isolating identified, small neurons from snail ganglia is described. The technique allows a particular neuron, previously identified by morphological and electrophysiological characteristics, to be marked and then isolated from the ganglia. This procedure was developed to permit the detailed comparison of the electrical characteristics of a neuron before and after isolation from an intact system. An earlier description has appeared is. The cell somata is marked intracellularly by the iontophoretic injection of Procion navy blue H3RS which visually differentiates the cell from other cells in the ganglion. The ganglion is then treated with a trypsin-haluronidase solution to soften the ganglion sheath, which is then removed. The cells are gently shaken to isolate them from the ganglion and then examined electrophysiologically. A comparison of membrane properties, such as action potential height, duration and rate of rise and decay was made before and after all treatments were applied to assess deleterious effect. An analysis of network properties, such as burst duration, number of spikes per burst and presynaptic activity was also performed after each phase of the procedure. No significant differences were noted after dye injection, enzyme treatment, and where appropriate, after isolation. An increase in input resistance and corresponding decrease in the slope of the steady state current-voltage plot (I-V plot) were observed after isolation of a cell. These were expected results of removing the 'load' (i.e. axon or electrical coupling) from the cell soma. This method may be applied to many other systems to study the effects of network interactions on the properties of a single cell and should therefore facilitate the analysis of neuronal networks as well as single cell properties.
220 INTRODUCTION It is becoming increasingly apparent that individual cell properties are important in determining the generation of activity by a network 23. It is therefore necessary to study a neuron in isolation in order to examine its specific membrane properties separately from those electrical and chemical synaptic interactions with other neurons. This paper describes a single cell isolation procedure applicable to small, identified neurons. Some separation of network properties from intrinsic properties has been accomplished using a variety of techniques - - most notably chemical and physical isolation. The use of EDTA is a commonly used chemical method of isolating cells from synaptic inputL However, the reduction of calcium in the solution as a result of the chelating action of EDTA can be expected to affect Ca-dependent membrane properties 6. Another method of isolation from synaptic input is replacement of Ca in the Ringer's solution with magnesium, which has the same problem. Physical isolation of neurons was first accomplished by Oomura and Maeno t,9. The isolated cells of this study exhibited a decrease in membrane potential and did not show spontaneously firing action potentials upon penetration with a microelectrode, indicating possible damage to the cells. Other isolation procedures have been suggested, but these are confined to studying the largest cells in the ganglion. Usually such techniques do not allow a small cell to be studied in the intact system and in isolation due to the difficulty of following the cell through the isolation procedure. Therefore, averaging of the parameters under investigation from many cells is often utilized instead of studying the same cell or cells before and after isolationl,3,4,9,1~, 17. This investigation described two related techniques: (1) a method for isolating a cell from a ganglion with minimal damage; and (2) a method for marking a particular cell, thus distinguishing it from the rest of the cells in the ganglion, allowing the cell to be identified after isolation. These two techniques permit the properties of the same cell to be studied while in a network of cells and after being removed. The experimental paradigm consisted of comparing various parameters, before and after isolation, which were chosen to reflect both the integrity of individual cells and synaptic interactions. The medial giant cells (cells 5) and the main pair of protractor neurons (cells 19) were the subjects of this study I1. The medial giant cells (75-100/~m in diameter) have little synaptic input, and are only distantly electrically coupled. The protractor neurons (approximately 50/~m in diameter) are members of a group of electrically coupled motoneurons which receive large, rhythmic IPSPs from the cyberchron network, a group of rhythmically bursting cells T M . The cyberchron cells are also an electrically coupled group of cells whose output drives the motoneurons (e.g. protractor motoneurons, such as 19 and retractor motoneurons) involved in the feeding behavior of the snail. Therefore, monitoring the large IPSPs on cells 19 also serves as an indicator of the integrity of the cyberchron network as well as that of cells 19.
221 MATERIALSAND METHODS
Specimen preparation Specimens of the fresh-water snail, Helisoma trivolvis albino, were maintained in a well-aerated aquarium and kept on a 12 h light/dark cycle. Details of animal care, dissection of the buccal mass (BM) and buccal ganglion (BG) and preparation immobilization for recording have been previously described in Kater et al. la,14. Cells were identified according to the system of Kater 14,15. All experiments and dissections were performed with the preparation immersed in physiological saline consisting of: 51.3 mM NaC1, 1.7 mM KCI, 4.1 mM CaCI2, 1.5 mM MgCI2, and 1.8 mM NaHCOa added as a buffer (pH 7.6) (see refs. 11-13).
Electrical recording Glass microelectrodes filled with 2 M potassium citrate (DC resistance = 20-40 Mf~) were used for intracellular recording, and an Ag-AgCI wire served as the indifferent electrode. A high input impedance preamplifier with a bridge circuit for injecting current (Getting Microelectrode Amplifier, Los Altos, Calif., Model 4) was used for intracellular recording and signals were displayed on a Tektronix 5000 series oscilloscope and on a Brush (Model 280) chart recorder. Action potentials were photographed from the oscilloscope using a Tektronix (Model C5) oscilloscope camera. Microelectrodes for dye injection were filled with a 4 % solution of Procion navy blue H3RS (PNB) (Polysciences), a negatively charged dye, in distilled water (DC resistance : 60-100 Mf~). PNB was injected into the cell by passing hyperpolarizing current of 1-10 nA through the electrode until a small area of blue appeared in the cell. The duration of current application varied with the electrode characteristics and the amount of applied current, but never exceeded 5 min. This blue area remained visible in the cell for at least 10 h. Cells marked in this manner were readily distinguishable from other cells in the ganglion even when all morphological characteristics (such as position and relative size of the cell in the ganglion) and characteristic electrical activity were removed by the cell's isolation from the ganglion. Cell membrane parameters were measured before and after each experimental manipulation to examine effects on the cell and included: (1) membrane potential; (2) voltage response to a step input current from which time constants, input resistance and steady-state I-V plots were determined; and (3) action potential characteristics, such as height, duration, and rates of rise and fall. Network interactions were assessed before and after each experimental manipulation by measuring: (1) feeding cycle time (defined as the number of bursts occurring in a given time recorded in protractor motoneuron 19); and (2) the shape of the IPSPs recorded in protractor motoneuron 19.
The isolation procedure The cell was marked with PNB after all initial measurements and experiments were performed. The BG and BM were then softened with the enzyme digestive
222 treatment which consisted of the following steps: (1) a 5 min soak in 0.12 % solution, by weight, of trypsin-hyaluronidase (Calbiochem) dissolved in saline, at 22--25 C , softening the sheath of the ganglion to facilitate the sheath's removal; and (2) two 10 min washes of saline at 5 °C. The BM and BG were either repinned in the dish used for electrical recording and data was again taken, or the isolation procedure continued with the BG being separated from the BM and the former placed in the cell isolation apparatus. The cell isolation apparatus consisted of a small petri dish whose bottom was covered with a layer of Sylgard, and then a layer of 0.5 % Agar (Difco, dissolved in saline). The cells clung to the surface of the Agar when they were isolated, immobilizing them and permitting penetration with a microelectrode. The petri dish was placed in a larger dish filled with ice which increased the number of viable cells and their lifespan. After pinning the BG, its sheath was cut away using electrolytically sharpened Tungsten needles. Gentle shaking of the ganglion isolated the cells from each other and from the neuropile. By varying the vigor of shaking, the amount of remaining axon could he controlled to some extent. The indifferent electrode was then placed in the dish containing the isolated cells, and standard electrical recordings (see Electrical recording) were taken. A cell which had been stained with PNB was easily visually differentiated from surrounding cells permitting comparative studies to be conducted. RESULTS
Effect of Procion navy blue A comparison of the activity of a cell before and after injection of PNB (cell 19; Fig. 1) shows very little change in action potential characteristics, and a small change in presynaptic activity and cycle time. There was an average change of 6.0 ~ for 10 preparations. Comparisons of the cell's input resistance and time constants (Fig. 1) before and after dye injection also shows little change. The membrane potential of the cell before intracellular marking was --60 mV, compared to --58 mV after, indicating B-PNB
,
,
i _
_
Im2v
J
0.5sot Fig. 1. Activity of cell 19 before and after PNB injection. B-PNB: before dye injection. AoPNB: after dye injection. Top traces show activity of the cell as a result of input from the network. Bottom traces show cell response to a small, hyperpolarizing step current injected.
223
CURRENT (nA)
0.8 0.6
•
0.4
•
0.2 "A I --30
I -20
I
I -10
10
-0.2
I 20
I
30
VOLTAGE
(mY)
-0.4 &Q -0.6
-0.8 Fig. 2. Steady-state 1-V plot of a cell (cell 19) before and after PNB injection. In all I-V plots shown, the plot was constructed by measuring the voltage response of the cell to different step current injections. A : before injection; • : after injection. a very small effect o f the dye. In 10 other cells examined, the m e m b r a n e potential never changed by more than ± 2 mV after dye injection (in 4 cases Em became more negative, in 5 cases more positive, and in 1 case remained unchanged). It was also noted that the cell could still fire out o f anode break after being marked with the dye, providing additional evidence that the cell was not affected by the dye injection. The steady-state I - V plot (Fig. 2) o f a cell (cell 19) before and after dye injection indicates that the steady-state m e m b r a n e properties were not affected by this
TH
~
~.'
Before
After
2O
ms
Fig. 3. Examination of the action potential of a cell before and after a given treatment. TH: before and after trypsin-hyaluronidase treatment (cell 19). PNB: before and after PNB injection. Top trace of each inset shows the voltage response of the cell. Bottom trace of each inset shows the current trace. The treatments were given separately to two diffrent cells.
224
I-
i
,J4omV 2sec
L___I
I nA 0.5 sec
Fig. 4. Cell 19 activity before and after trypsin-hyaluronidase treatment. B-TH: before application of the enzymes. A-TH: after application of the enzymes and washes. Top traces show activity of the cell as a result of input from the rest of the network. Bottom traces show cell response to a small, hyperpolarizing step current injected.
treatment. The passive membrane region appears to be linear before and after treatment. A closer look at an action potential of a marked cell (cell 19, Fig. 3) indicates that action potential duration, threshold and height were not significantly altered. The above tests demonstrate that the passive membrane properties and those active membrane properties contributing to action potential generation do not appear to be affected by PNB injection.
CURRENT (r~A) L0'
t
0.8" 0.6'
•
0.4" 0.2-
-30
-16 •'
-2o
-0.2•
•
*
O
1'o
2b
3o
VOLTAGE (mY)
-0.4"
-0.6.
Fig. 5. Steady-state I-V plot of a ~11 (cell 5) before and after trypsin-hyaluronidase treatment. ~ , before treatment; O, after treatment. For an explanation of how points were obtained, see Fig. 3 legend.
225
Effect of trypsin-hyaluronidase treatment The same parameters used to analyze the extent of damage in the PNB-injected cells were used to assess the effects of the enzymatic treatment, i.e those of network interaction and of cell membrane response. Action potential characteristics, synaptic activity, and IPSP frequency were not significantly affected by the enzyme treatment (cell 19, Fig. 4). A comparison of passive membrane properties also showed that input resistance and time constant were not noticeably affected. The passive membrane region (where I-V plot is linear) and region of rectification exhibited little change after treatment (cell 5, Fig. 5). Comparison of action potential characteristics (e.g. height, duration, and rates of rise and decay) (Fig. 3) before and after the enzymatic treatment provides further evidence that little alteration has occurred in the cells as a result of this treatment. The resting potential of only one cell out of ten different cells examined was slightly affected (more positive by 3 mV). Delayed and combined treatment effects Possible deleterious effects due to the combination of both PNB injection and enzymatic digestion were investigated. The cell was examined at 2, 4, 6 and 8 h after application of both treatments in order to demonstrate possible long term effects. A comparison of the I-V plots of the cell before and 8 h after administration of the treatments showed insignificant differences in both the rectification and the passive linear regions of the steady-state I-V plots (Fig. 6). The isolated cell After analyzing and marking a cell and enzymatically treating the BG and BM the cell was physically isolated from the rest of the cells in the ganglion. Potential deleterious effects from the isolating procedure were assessed using both morphological observations and electrical measurements. As described in Methods (the isolation procedure), cells were isolated by gently shaking the slit ganglion in fresh saline.
CURRENT • 0.3~A) 0'2! 0.1t . :zG -,'s -lb -;, I[ ae
•A A •&
•
$
-0.1
-0.2 &
-0.3 Fig. 6. Steady-state I - V plot of a cell (cell 5) before both a PNB injection and a T H treatment were administered to the same cell, and 8 h after both treatments were administered. A , before treatment; Q, after treatment. For an explanation of how points were obtained for this plot, see the legend of Fig. 3.
226
Fig. 7. A series of pictures of 4 different ganglia taken at various stages of the isolation procedure. A : two different cells have been injected with PNB. Injected cells are marked with arrows. Note how the dye is contained inside the ceil. Dye stays visible for a minimum of 8 h after injection. B: the ganglion has been separated from the rest of the preparation, has been cut circumferentialty, and the cells have started to emerge. C: the sheath has been lifted away from the main body of cells, and the ganglion has been lightly shaken. Note that several cells are already completely isolated. D: a completely isolated cell. Bits of glia are still clinging to the axon as evidence of the gentleness of the techniuqe. Also note that the membrane integrity has been maintained, and that the nucleus is small and round, more evidence that this cell is healthy. All calibration marks equal to 100/ml.
M o r p h o l o g i c a l e x a m i n a t i o n of an isolated cell (Fig. 7) d e m o n s t r a t e d the gentleness of this technique in r e m o v i n g a cell from the ganglion. The cell retained a length of axon over 100 # m long, a n d on the axon clinging fragments o f glial cells were visible. The nucleus of the cell was r o u n d a n d small, similar to its a p p e a r a n c e in an unisolated cell. Lastly, the m e m b r a n e of the soma appeared to have m a i n t a i n e d its integrity with n o a p p a r e n t or visible 'blebs' ( m e m b r a n e outpocketings) due to m e m b r a n e degeneration. Inspection o f an action potential (cell 19, inset, Fig. 8) of a cell before a n d after its removal from the ganglion shows insignificant effects on height, d u r a t i o n , rates of rise a n d decay. The I - V curves of this cell before a n d after
CUI~nR~NT 1.0. ,:
O.8
B-I
0.6-
',.____
"~
•
0.4-
A-I
0.2-
-~'o
•
•J
-2b -10**•
1'0
2'0
3"0 4"0 VOLTAGE
-~2
•
-0.4 -0.6
Fig. 8. Steady-state I - V plot of a cell (cell 19) before and after isolation. Insets show the action potential of the cell before and after isolation• Top trace of the insets is the voltage response of the ceil. Bottom trace of the insets shows the current injected. A , before isolation; O, after isolation•
• O.E
~....
•
A-I
"-~
--~
~'...
",, ' "" >"I 0.(
", -...'%,
"...-
-0~
* ---l %.."~Q ~.."%,
"A''i B-I
--I.
7
~
l.'v '
V-- 10.2 ~lnll l
i""e'-,~
.~----
-,,,
"z.,, "-*,
-L6 ---I
1'o
'
J
3'0
",.,..
'
sb
TIME (msec)
'
7'0
'
9'0
Fig• 9. Comparison of time constants before and after isolation of cell 19. This graph illustrates that the non-isolated cell response to a small hyperpolarizing current pulse has two components: the fast component with the steep slope is due to the axon; and a slow, gradual sloped component (due to the cell soma)• The isolated cell has only one component - - the slow soma response. The insets show the response of the same cell before and after isolation to the same size current injection• B-I, before isolation; A-I, after isolation. The top trace of each inset shows the membrane voltage response and the bottom trace is the current injected.
228 isolation, however, showed a change in the slope of the passive linear region of the celt. The input resistance increased as expected with isolation of the cell. The two regions of rectification (hyperpolarizing and depolarizing) were rotated due to the general increase in steady-state resistance. A more detailed look at the passive membrane properties was obtained by examining a graph o f L n (Vr-V(t)) vs time (cell 19, Fig. 9). Various linear regions of this graph correspond to different time-constant processes. Two time-constant processes were observed from the intact cell in the ganglion: (1) a slow time-constant (90 msec) corresponding to the soma time constant; and (2) a faster time-constant (33 msec) due to the properties of the axon 7. In the isolated cell, one time-constant was seen. The faster time-constant was abolished as a result of removing most of the axon. DISCUSSION Procion navy blue injection As has been previously described, a cell in a ganglion can be identified by its characteristic electrical activity, its relative morphological location and its size. However, when a cell is isolated these criteria can no longer be applied to most cells, and a direct comparison of a cell's properties before and after isolation cannot be made. A way to overcome this difficulty is to physically mark the cell to be studied. This study utilized the dye Procion navy blue H3RS which we have shown to have insignificant effects on several membrane properties. However, there has been some evidence reported that the Procion dyes affect the membrane properties of the cell, and also the cell's internal structuresS, 14,2°. There are several possible explanations for these discrepancies. In earlier studies the dye was injected for several hours, increasing the amount of dye injected and applied current. Enzymatic treatment Many studies have demonstrated the deleterious effects of digestive enzymes on membrane properties3,6,2L An increase in action potential duration, membrane depolarization and inactivation of specific currents after prolonged exposure to digestive enzymes such as trypsin, hyaluronidase and pronase has been reported 6,~z. In this study, by carefully modulating solution temperature and the application time of the enzymes, the damaging effects of the enzymes were minimized. No significant differences were seen in individual cell parameters and synaptic interactions between neurons. It is important to note that the exposure time of the preparation to the enzymes was considerably shorter than that of previous studies ~,6,22 (5 min compared to several h). In this study the enzymes were used to soften, but not dissolve the sheath, allowing it to be cut easily without mechanically stressing the cells in the ganglion, other studies applied the enzymes to the ganglia for lengths of time causing the sheath to fall apart 16. The isolation This study makes a direct comparison of the membrane and network properties
229 of a cell that is still connected to other cells in the ganglion and the same cell after being physically isolated from other cells. The isolation technique was shown to be very gentle by use of morphological evidence (Fig. 8) and by comparisons of some relevant membrane and network properties, action potential characteristics, and the rectification areas of the I - V plot, which appear to be unaffected by the isolation procedure. Previous methods have employed large cells and have used the average value from several cells to determine the effects of isolating a celp6,17. Several of these studies used cells that exhibited some degree of mechanical damage, produced either by cutting a sheath that was not softened or by micropipetting the cells from the neuropile during the isolation process16, is. In the present series of experiments it was noted that pipetting the cells resulted in some damage to the cell, and decreased the number of viable ceils, Gently shaking the ganglion to free the cells prevents the stress of the cells bumping into the glass and the turbulence when drawn into the pipette. The effects of loading of the network were demonstrated by comparing the voltage response before and after isolation. It was seen that the axon of the cell can cause at least a 3-fold decrease in input resistance. This decrease was also seen in the change of slope of the passive linear region of the I - V plot. This change in input resistance was compared to the expected value based on the measured time constants (soma and axon) of the intact neuron and from a simple equivalent model 7. A 5-fold increase in resistance is predicted from the model, which is compatible with the 3-fold increase observed. The fact that the calculated increase based on the model is slightly greater than the observed increase is probably due to the fact that a small length of axon remained after isolation. This isolation technique allows a direct comparison of membrane and network parameters of a specific cell while the cell is still in the network, and once it is removed, the interactions of a cell with its surroundings can then be directly evaluated.
REFERENCES 1 Alving, B. O., Spontaneous activity in isolated somata of Aplysia pacemaker neurons, J. gen. Physiol., 45 (1968) 29-45. 2 Asada, Y. and Bennett, M. V. L., Experimental alteration of coupling resistance at an electrotonic synapse, J. Cell BioL, 49 (1971) 159-183. 3 Chen, C. F., von Baumgarten, R. and Takeda, R., Pacemaker properties of completely isolated neurones in Aplysia californica, Nature New Biology, 233 (1971) 27-29. 4 Faber, D. S. and Klee, M. R., Membrane characteristics of bursting pacemaker neurones in Aplysia, Nature New Biology, 240 (1972) 29-31. 5 Getting, P. A., Modification of neuron properties by electrotonic synapses. I. Input resistance, time constant and integration, J. Neurophysiol., 37 (1974) 846-857. 6 Heyer, C. B. and Lux, H. D., Properties of a facilitating calcium current in pace-maker neurones of the snail, Helixpomatia, J. Physiol. (Lond.), 262 (1976) 319-348. 7 Jack, J. B., Noble, D. and Tsein, R. W., Electric Current Flow in Excitable Cells, Clarendon Press, Oxford, 1975, pp. 160-172. 8 Jankowska, E. and Lindstrom, S., Morphological identification of physiologically defined neurones in the cat spinal cord, Brain Research, 20 (1970) 323-326. 9 Kado, R. T., Aplysia giant cells: soma-axon voltage clamp current differences, Science, 182 (1973) 843-845.
230 10 Kaneko, C. R. S., Merickel, M. and Kater, S. B., Centrally programmed feeding in Helisoma controlled by electrically coupled network. 1. Premotor neuron identification and characteristics, Brain Research, in press. 11 Kater, S. B., Feeding in Helisoma trivolvis: the morphological and physiological bases of a fixed action pattern, Amer. ZooL, 14 (1974) 1017-1036. 12 Kater, S. B., Heyer, C. and Hegmann, J. P., Neuromuscular transmission in the gastropod mollusc Helisoma trivolvis: identification of motoneurons, Z. vergl. Physiol., 74 (1971) 127-139. 13 Kater, S. B. and Kaneko, C. R. S., An endogeneously bursting neuron in the gastropod mollusc, Helisoma trivolvis: characterization in vivo, J. comp. Physiol., 79 (1972) 1-14. 14 Kater, S. B. and Nicholson, C., Intracellular Staining in Neurobiology, Springer-Verlag, New York, 1973, pp. 211-226. 15 Kater, S. B. and Rowell, C. H. F., Integration of sensory and centrally programmed components in generation of cyclical feeding activity of Helisoma trivolvis, J. Neurophysiol., 37 (1973) 142--155. 16 Kostenko, M. A., Geletyuk, V. I. and Veprintsev, B. N., Completely isolated neurons in the mollusc Lymnaea stagnalis. A new objective for nerve cell biology investigation, Comp. Biochem. Physiol., 49 (1974) 89-100. 17 Kostyuk, P. G., Krishtal, O. A. and Pidoplichko, V. 1[.,Effect of internal fluoride and phosphate on membrane currents during intracellular dialysis of nerve cells, Nature (Lond.), 257 (1975) 691-693. 18 London, J. A. and Merickel, M. B., Analysis of neural network components using single cell isolation, Neurosci. Abstr., II1,383. 19 Oomura, Y. and Maeno, T., Does the neurone somata actually generate action potentials?, Nature (Lond.), 197 (1963) 358 359. 20 Payton, B. W., Histological staining properties of Procion yellow M4RS, J. Cell BioL, 45 (1970) 659-662. 21 Purves, D. and McMahan, U. J., The distribution of synapses on a physiologically identified motor neuron in the central nervous system of the leech, J. CellBiol., 55 (1972) 205-220. 22 Rojas, E. and Armstrong, C. M., Sodium conductance activation without inactivation in pronaseperfused axons, Nature New Biology, 229 (1971) 177-178. 23 Russell, D. F. and Hartline, D. K., Bursting neural networks: a reexamination, Science, 200 (1978) 453456.