Gap junctions at electrotonic synapses formed by club endings on the Mauthner cell lateral dendrite in the goldfish. A: in the central region the unit membranes approach each other closely, but over much of the region of apposition a gap of about 20 Å remains. B: in this plane of section, spots of increased density appear in the centers of the apposed membranes along much of the junction. These spots have a periodicity of about 90 Å and are in register in the two membranes. C: lanthanum injected intraventricularly fills the extracellular spaces at either end of the gap junction and penetrates the gap to the center of the junction. The staining in the gap is interrupted, perhaps periodically. D: in tangential section, lanthanum staining in a gap junction appears as a more‐or‐less hexagonal lattice. Center‐to‐center spacing is about 90 Å. A few of the clear regions outlined by the hexagonal lattice have a central dense spot (arrow). Vertical bar, 500 Å for A, C, and D and 600 Å for B.

Electron micrographs of tight junctions between endothelial cells of cerebral capillaries in the mouse. In A‐C the vessel lumen is at the top and the basal lamina runs along the bottom. A: unit membranes contact each other and occlude the intercellular cleft at 3 points (arrow indicates the uppermost point of occlusion). Cleft is open the remainder of the distance to the basal lamina. B: intravascularly injected lanthanum fills the vessel lumen and the intercellular cleft down to a point of contact between the endothelial cell membranes. C: intraventricularly injected horseradish peroxidase permeates the basal lamina and passes up the intercellular cleft as far as a point of membrane contact. Aldehyde fixation. Vertical bars, 500 Å.

Gap junction as seen in freeze‐fracture preparation from mouse liver. At lower left the fracture plane reveals the cytoplasmic leaflet (A face) of the cell that lay away from the viewer; at upper right the outer leaflet (B face) of the near cell is revealed. In the center is the gap junction, which appears as a large aggregate of particles in the cytoplasmic leaflet and as complementary pits in the outer leaflet. In some small regions the particles form a regular hexagonal array, but the domains of close packing are not large. At large ridges in the top center and bottom right the fracture plane broke across the extracellular space between the 2 cells. The ridge in the gap junction between leaflets showing particles and pits is much lower and reflects the close apposition of the cell membranes in this region. As is commonly but not universally found in cells, the cytoplasmic leaflet of the nonjunctional membrane contains dispersed particles that are much more numerous than those in the outer leaflet. Inset shows a region of the cytoplasmic leaflet of the junction at higher magnification. Some of the particles appear to have a small pit at their centers. Vertical bar, 0.1 μm; 0.05 μm for inset

Freeze‐fracture preparation of gap junctions between neurons. The neuron is a spinal electromotor neuron of the gymnotid electric fish Sternarchus. In sectioned material only axosomatic gap junctions are seen 252. There are 4 aggregates of particles typical of gap junctions in other structures (gj 1–4). The somatic location of the junctions is indicated by the nucleus (Nc) and the nuclear pores (p), which were exposed when the fracture plane left the surface membrane to cross the cytoplasm (Cy). The exposed surface of the neuron is the cytoplasmic leaflet (A face). Aldehyde‐fixed tissue.

Occluding junction as seen in a freeze‐fracture preparation from epithelial cells of the rat small intestine. Microvilli protruding into the lumen point downward. The occluding junction runs along the center and appears as a meshwork of ridges and grooves. The surface at top is the outer leaflet (B face) of the cell that lay toward the reader. This leaflet contains the grooves of the fractured occluding junction. In the center and to the right of the junction are regions where the fracture plane broke through to the underlying cell exposing the cytoplasmic leaflet (B face), which contains the ridges. The membrane leaflets bulge toward the viewer between the grooves in the outer faces and away from the viewer between the ridges of the cytoplasmic faces; these bulges reflect the separation of the apposed membranes between the lines of contact. Vertical bar, 0.1 μm.

Effects of fixation on the appearance of gap junctions in thin section. All junctions are between ependymal cells near their apices. A: osmic acid fixation followed by uranyl acetate before dehydration (en bloc staining); the characteristic gap is seen to separate the 2 unit membranes. B: permanganate fixation; there is a central dark line in the junction, but no gap is apparent. C: osmic acid fixation, but no en bloc uranyl staining. The outer lamella of the unit membranes is unstained. There is no central staining in some regions of the junction. In other regions there is a periodic row of dots with a separation of ca. 100 Å (see also Fig. 24). Vertical bar, 500 Å

Gap junctional membranes isolated from rat liver negatively stained with phosphotungstic acid. In many regions, the junctional fragments show a regular hexagonal lattice that corresponds to the lattice seen in thin sections of lanthanum preparations. A dense spot is present in the centers of the clear areas outlined by the lattice. The periodicity of the lattice is about 100 Å. Vertical bar, 500 Å.

Intercellular passage of the dye Procion yellow between axonal segments of crayfish septate axon. Dye was injected into the posterior segment (Ap) through a recording microelectrode that allowed monitoring of axonal impulses. The dye passed into the anterior segment (A_p) across the septum, which runs between arrows. Dye was visible in the anterior segment within 1 h, but the preparation was allowed to stand for about 15 h before it was photographed through a fluorescent microscope. (The lower molecular weight and more fluorescent dye fluorescein visibly crosses the septum within a few minutes.) Some axoplasm had coagulated in the posterior segment and was more brightly fluorescent. Ganglion, connectives, and peripheral nerves are faintly discernible through their autofluorescence. Maximum axon diameter is about 150 μm. The anterior axon narrows sharply caudal to the septum and can be seen to run ventromedially, which is toward its soma on the contralateral side

Diagrammatic representation of a small area of gap junction. Two apposed membranes are shown in perspective as in the horizontal plane. At right the junction is split and the upper membrane is peeled back to reveal the external aspect of the lower membrane. The gap between the apposed membranes and the structures that bridge it are indicated. The membranes each contain subunits that form hemichannels and correspond to the particles seen in freeze‐fracture. Hemichannels are shown as separated by lipid in the plane of the membrane. Two aligned hemichannels form a complete channel connecting the cells interiors. Separated hemichannels are probably not open to the exterior, but no gating mechanism is shown.

Electrical model for an electrotonic synapse. A: 2 cells that would ordinarily be effectively isopotential. Voltage and current electrodes are indicated in each cell with measured voltages (V₁ and V₂) and applied currents (i₁ and i₂). The indifferent, or ground, electrode in the bathing medium is omitted. B: an equivalent circuit of the cells. C₁ and C₂ are cell capacities; r₁, r₂, and r_c are cell and junctional (or coupling) resistances. Resistances of cytoplasm and external medium are neglected. Junctional capacity is generally negligible and is omitted.

Electrotonic coupling of supramedullary neurons of the puffer fish. Upper trace, polarizing current. Middle and lower traces, intracellular recordings from adjacent neurons, cells 1 and 2, respectively. A: multispike response to spinal stimulation. The same number of responses occur in each cell, although the somas are not invaded during the 2nd and 3rd responses and only axon spikes are recorded. B: a depolarizing pulse which is adequate to evoke a spike, is applied to cell 2. A somewhat delayed spike is observed in cell 1, indicating propagation across the electrotonic synapses. There is also appreciable electrotonic spread of the maintained depolarization. Higher gain recording in cell 1. C and D: electrotonic spread of hyperpolarization when current is passed in cells 1 and 2, respectively. Higher gain recording in the unpolarized cell. The electrotonically spread potentials are attenuated and slowed.

Current‐voltage relations for electrotonic coupling of the cells in Fig. 11. Cell 1 and Cell 2 correspond to the cells of the middle and lower traces, respectively. Hyperpolarization in the lower left quadrant. Larger units on the ordinate are for potentials in the polarized cell (•); smaller units are for potentials in the other cell (×). Input resistances (r₁₁ and r₂₂) of the 2 cells are linear for moderate hyperpolarization and differ slightly. Over the same range, the transfer resistances, that is, the ratio of potential in one cell resulting from current in the other (r₁₂ and r₂₁), are approximately equal, and the same resistance line is drawn for both sets of data. For larger hyperpolarizations and for depolarizations, input resistances become nonlinear, but the amount of electrotonic spread is not changed in proportion to change in resistance.

Equivalent circuits of a “leaky” junction. A: circuit where the current path out of the middle of the junction into the extracellular medium is represented by a resistance. B: circuit equivalent to that in A obtained by a π‐T transformation of the central 3 resistors. The “leak” is electrically equivalent to additional resistances in parallel with the pre‐ and postjunctional resistances. Abbreviations as in Fig. 10.

Calculated PSP's for an impulse into the model electrotonic synapse (Fig. 10B). The input function is V₁ = (t/A)² exp (−2t/A + 2), where t is time and A is time to peak; its impulselike shape is indicated by the curve labeled x = 0; x is the ratio of the postsynaptic time constant (including junctional resistance) to the time to peak of the impulse. The amplitude of the PSP (V₂) normalized by the steady‐state coupling coefficient k from V₁ to V₂ is plotted against time measured in terms of rise times of the impulse, t/A. If the rise time of the presynaptic impulse is 1 ms and the postsynaptic time constant is 6.4 ms, x is 6.4 and the observed delay is about 0.3 ms. If the rise time is 0.5 ms, x is 12.8 and the delay is about 0.4 times the rise time or 0.2 ms.

Delays at electrotonic synapses. A: transmission across the septum of the septate axon; R and C indicate impulses in the rostral and caudal segments. Propagation is in the rostrocaudal direction. Delay is less than 0.05 ms. B: transmission from giant fiber to motoneuron in the hatchetfish; the delay is about 0.05 ms; several superimposed sweeps. [From Auerbach & Bennett 9.] C: sonic muscle motoneurons in the toadfish, delays in transmission between motoneurons. Threshold stimulation of the axon of the penetrated cell, about 10 superimposed sweeps. The middle trace is the intracellular record. The upper trace is a record of the antidromic volley at the edge of the spinal cord. The lower trace is recorded later just outside the cell at the same stimulus strength and should be subtracted from the intracellular record to give transmembrane potentials. When the antidromic spike is evoked, the potential rises rapidly off screen. When the axon is not stimulated a positive negative field potential preceeds a slower depolarization of the cell membrane. This depolarization occurs at a latency of about 0.3 ms from the shortest latency antidromic spike and from the peak of the extracellular positivity generated by excitation of the neighboring cells. D: as in C but recorded at a slower sweep speed at different stimulus strengths. The antidromic depolarization is graded with the magnitude of the antidromic volley. [From Bennett 18.] E: delays in transmission between supramedullary neurons. Threshold stimulation of one cell shows that the postsynaptic component associated with the spike is delayed by about 2 ms. Voltage calibrations: 50 and 5 mV for high and low gain respectively. Time calibrations: 1 ms for A, C, and D; 0.2 ms for B; and 5 ms for E. [From Bennett et al. 35.]

Transmission of impulses across a rectifying electronic synapse, the giant motor synapse of the crayfish. A: an antidromic impulse in the postsynaptic (post) motor axon causes only a very small potential in the presynaptic (pre) fiber. B: a directly evoked impulse in the presynaptic fiber causes a large PSP in the motor axon.

Effect of hyperpolarization on the PSP at a rectifying electrotonic synapse, the giant fiber‐motoneuron synapse of the hatchetfish. Inset: upper trace, polarizing current; lower trace, recording in a motoneuron. Hyperpolarization augmented the PSP evoked by spinal stimulation, which activates the giant fibers. There is little change in PSP shape but a doubling in amplitude (3 superimposed records). Graph, relation of PSP amplitude and hyperpolarization; data from the same experiment as the inset.

Electrically mediated inhibition at the Mauthner cell initial segment. A: recording inside and just outside the initial segment (inset).A positivity (arrow) is recorded outside about 1 ms after the large negative potential generated by the Mauthner cell axon; at the same time only a small positivity is recorded internally. Subtraction of the records (black line on white) gives the transmembrane potential; the membrane is hyperpolarized during the external positivity. The later long‐lasting positivity recorded intracellularly is an IPSP that is inverted because of leak of Cl from the recording electrode. B: stimulation of one of the electrical inhibitory neurons (upper trace) generates a small extracellular positivity of ca. 0.4 mV in the initial segment region of the Mauthner cell (lower trace). C: activation of the Mauthner cell by spinal stimulation produces a negative potential in an inhibitory neuron (upper trace), which has the same latency and time course as the negativity outside the initial segment of the Mauthner cell, although it is smaller. This response is much smaller just outside the inhibitory interneuron (lower trace), so that the membrane is hyperpolarized. D: this hyperpolarization decreases cell excitability to direct stimulation (superimposed traces with and without Mauthner cell stimulation, which is indicated by •). Time calibration same for C and D. [B‐D from Korn & Faber 188.]

Diagram and equivalent circuits of the Mauthner cell electrical inhibitory synapse. A: relation between the cells and lines of current flow are indicated on an equivalent circuit. Voltage generators are shown in the initial segment region of the Mauthner (M) cell and in the soma of the inhibitory (I) cell. The resistance of the I cell terminal is r₁. The resistance of the M cell soma is r_s. The extracellular (access) resistance close around the initial segment and inhibitory terminal is r_e. The low resistance of the large volume of tissue farther away is neglected. B: the equivalent circuit rearranged to emphasize its symmetry. Activity of the I cell produces current from right to left through r_e; an extracellular positivity is generated that hyperpolarizes the M cell initial segment. Activity of the M cell causes current to flow from left to right through r_e; the extracellular space outside the I cell soma is positive compared to that outside its terminal and the soma membrane is hyperpolarized

Electrical coupling between cells that have a circular region of low‐resistance membranes closely apposed. Left, potential in the gap V_g on the right. Because the gap widens at the edge of the apposition, V_g is assumed to go to zero at this point. V₁ and V₂ are the potentials in the pre‐ and postappositional cells, respectively. Current is applied in the pre cell setting up V₁; the input resistance of nonappositional membrane in the post cell is assumed to be high relative to the resistance of the appositional membrane. With these assumptions and an apposition with a radius of 4 space constants defined in the usual way, the curved line in the graph shows V_g as a function of radial position in the gap. This could occur for appositional membrane with a resistivity of 1 Ωcm², a gap width of 200 Å, a gap resistivity of 100 Ωcm, and an apposition of 2.8 μm in diameter. The coupling coefficient, V₂/V₁ is approximately one‐third.

Properties of the giant electromotor neurons of the electric catfish. A: upper and lower traces, recordings from right and left cells, respectively. Brief stimuli of gradually increasing strength are applied to the nearby medulla (several superimposed sweeps; the stimulus artifact occurs near the beginning of the sweep). Depolarizations of successively increasing amplitude are evoked until in one sweep both cells generate spikes. B: 2 electrodes in the right cell, one for passing current (shown on the upper trace) and one for recording; one recording electrode in the left cell. The lower traces, which are from the recording electrodes, start from the same base line. When an impulse is evoked in the right cell by a depolarizing pulse, the left cell also generates a spike after a short delay. C: when a hyperpolarizing current is passed in the right cell, the left cell also becomes hyperpolarized, but more slowly and to a lesser degree (display as in B). D: when organ discharge is evoked by irritating the skin, a depolarization gradually rises up to the threshold of the giant cell and initiates a burst of 3 spikes (lower traces, base line indicated by superimposed sweeps). Each spike produces a response in the organ (upper trace, recorded at high gain and greatly reduced in amplitude because curare, used to prevent movement, also blocked transmission from nerve to electrocyte)

Effect of polarization in a single pacemaker cell of the weakly electric gymnotid Gymnotus. Upper trace, activity in the spinal cord and peripheral nerves leading to the electric organ (recorded by needle electrodes at high gain in a curarized animal); middle trace, intracellular recording from a pacemaker cell; lower trace, current applied through the recording electrode. Two superimposed sweeps in each record, one with and one without applied current. The sweeps are triggered by the spike of the pacemaker cell. Faster sweep in A and B. A and C: a depolarizing pulse that evoked a spike advances the next and subsequent spikes but does not desynchronize or itself cause any descending activity. B and D; a hyerpolarizing pulse retards the next and subsequent spikes but does not desynchronize the descending activity.

Responses of pacemaker and relay cells in a weakly electric gymnotid Gymnotus. Upper traces, activity in the spinal cord and peripheral nerves leading to the electric organ (recorded by needle electrodes at high gain in a curarized animal); lower traces, intracellular recordings in pacemaker (A and C) and relay (B and D) neurons. Faster sweep in C and D where dotted lines indicate the times of firing of the cells in relation to the descending activity.

Electronic coupling by way of dendrodendritic synapses in electromotor relay nuclei of a mormyrid. A: light micrograph of silver‐stained preparation (Romanes' method) of the medullary relay nucleus. A thick bridge appears to connect the 2 cell bodies without an obvious intervening membrane. B: with electron microscopy such bridges are seen to have partitioning membranes (between arrows) that separate the cell bodies (s). Axon terminals (a) occur on the cells and myelinated fibers and blood vessels (bv) are present. C: at a similar region of apposition between spinal relay neurons, higher magnification shows large regions where the membranes are very closely apposed. The membranes show the ordinary extracellular space at the top and bottom of the micrograph. Material was fixed in osmic acid, and sections were stained with uranyl acetate and lead citrate. The central region of the junctions shows dots that may be periodically arranged in the lower part of the figure (see Fig. 6).

Electrotonic coupling by way of presynaptic fibers in the spinal electromotor nucleus of the electric eel. A: an axon (A) appears to synapse with the soma of an electromotor neuron (S2) and with a short dendrite from another cell (SI). B: higher magnification from the same section showing a close apposition (probable gap junction) between A and the dendrite of S1. C: a neighboring section showing a close apposition between A and S2. The axon terminal provides a short pathway between cells that can be expected to contribute to the coupling observed physiologically (dendrodendritic electrotonic synapses are also present in this nucleus). Some presynaptic vesicles are present, although no chemically mediated component is seen in the PSP's.

Synaptic control of electrotonic coupling: neurons controlling pharyngeal expansion in the mollusc Navanax. Upper traces, recording from an M (medium‐sized) cell. Middle traces, recording from an ipsilateral G (giant) cell (higher gain in row of B, B′, D, D′). Lower traces, polarizing current applied to G cell for upper records and in M cell for lower records. Primed letters indicate that a decoupling train of stimuli is applied to the large ipsilateral pharyngeal nerve at the beginning of the sweep. A and B: electrotonic spread of hyperpolarization. C and D: spread of depolarization. Spikes in the polarized cells produce small, somewhat slowed, depolarizing components in the unpolarized cell. These relatively brief potentials are superimposed on a slow depolarization. Irregularity of firing in C is due to activation of an inhibitory interneuron by stimulation in the G cell. A′ and B′: electrotonic spread of hyperpolarization is practically eliminated after decoupling stimuli. Input resistance of the cells is reduced to about one‐half. During hyperpolarization small oscillations become visible which represent variations in the synaptic activity responsible for the decoupling. These oscillations are much smaller at the resting potential because their reversal potential is near this level. C′ and D′: electrotonic spread of depolarization is greatly reduced after decoupling stimuli. Reduction in spread of maintained depolarization is greater than reduction in spread of spikes. Stimulation in the G cell no longer can excite the M cell. The M cell can still be excited by stimuli applied in its own soma, although its excitability is reduced (D′).

Diagram and equivalent circuits for synaptic control of electrotonic coupling between neurons. A: simplest equivalent. Each cell and the coupling resistance are represented by a single resistor (r₁, r₂, and r_e, respectively). Decrease in r₁ or r₂ by inhibition would decrease coupling, as well as inhibit the cells. B and C: strategic localization of inhibitory synapses along the coupling pathway (filled circle endings in C) could allow effective decoupling without a great degree of inhibition of firing evoked by other inputs. Additional resistances in the equivalent circuit are r_i for the inhibitory synapses and r_s for the series resistance of the collaterals forming the coupling pathway. As r_i becomes zero the coupling between cells vanishes, but excitatory synapses located closer to the soma could depolarize r₁ and r₂ and cause the cells to fire independently.

Diagram of somatic and dendritic inputs to medial rectus oculomotor neurons of a teleost. Dendritic inputs (left arrow) are activated by stimulation of the ipsilateral eighth nerve. There is no coupling, and movements are smoothly graded in amplitude. The somatic inputs (right arrow) are activated by stimulation of the ophthalmic nerve or contralateral eighth nerve. There is weak coupling between the cell bodies (by way of the presynaptic fibers) and some increase in synchronization of firing results.

Impulses arising from dendritic or somatic depolarizations in oculomotor neurons of a teleost. Upper and middle traces, intracellular recording at high and low gain from a medial rectus oculomotor neuron. Lower traces, efferent activity in the medial rectus nerve. A₁ and A₂: threshold stimulation of the ipsilateral eighth nerve, dendritic impulse initiation. Little PSP is seen whether or not an abruptly rising spike occurs. A₃: stronger stimulation evokes a multispike discharge. The first 2 spikes arise abruptly. The last 2 are preceded by some PSP. B₁ and B₂: threshold stimulation of the ipsilateral ophthalmic nerve; impulses arise near the soma. Large slow PSP's appear to initiate the spike. B₃: stronger stimulation evokes a more rapidly rising PSP and multiple spikes.

Differential effects of hyperpolarizing current on impulses arising in the dendrites and close to the soma. Upper trace, intracellular recording from a medial rectus oculomotor neuron. Middle trace, current passed through the recording microelec‐trode with a bridge circuit. Lower trace, efferent activity in the medial rectus nerve. A₁: spikes initiated in the dendrite arise abruptly from a level base line in response to stimulation of the ipsilateral eighth nerve. A₂: when the same stimulus is given during a hyperpolarizing current pulse, the first response is delayed and the number of spikes is reduced, but little PSP is recorded at the times that the first and second spikes arise in A₁ (2 superimposed sweeps with and without nerve stimulation). B₁: spikes initiated near the soma are preceded by an obvious PSP that is evoked by ipsilateral ophthalmic nerve stimulation. B₂: when the same stimulus is given during a hyperpolarizing pulse, spikes are blocked revealing a large PSP (2 superimposed sweeps, with and without nerve stimulation).

Effects of polarizing currents on EPSP's associated with conductance increases and decreases. A: the fast EPSP evoked by a single stimulus to the presynaptic nerve is associated with a conductance increase. Hyperpolarizing (negative) currents augments the EPSP. Depolarizing (positive) currents decrease and then invert the EPSP. B: the slow EPSP is associated with a conductance decrease. After block of the fast EPSP with nicotine the slow EPSP is evoked by presynaptic stimulation at 100/s for 2 s (indicated under lowest trace). Although the slow EPSP is a depolarizing response, it is augmented by depolarization and decreased by hyperpolarization. The reversal potential for the slow EPSP is close to that for the after hyperpolarization of the spike, which suggests that the slow EPSP is due to a decrease in K conductance.

Dual‐action excitatory‐inhibitory PSP's in the buccal ganglia of Aplysia. The pre‐ and postsynaptic elements are subscripted 4 and 7, respectively. B, buccal; L, left; R, right. A: effects of altering the postsynaptic membrane potential on the PSP's and on the potentials produced by iontophoretic application of ACh to the cell soma. At the resting potential (0, middle traces) PSP shows a slight negative phase and the ACh potential is monophasic. When BL₇ is depolarized by 20 mV (upper traces) PSP develops a pronounced hyperpolarizing phase after the initial depolarizing phase, and the ACh potential is also diphasic. When BL₇ is hyperpolarized by 20 mV (lower traces) both PSP and ACh potential are monophasic and depolarizing. B: postsynaptic changes in excitability caused by the 2‐component PSP's. 1, when BR₇ is at its resting potential, activation of BR₄ causes excitation that augments during a high‐frequency burst; 2, when BR₇ is depolarized to fire at a slow rate, activation of BR₄ causes first inhibition and then excitation. C: pharmacological separation of the 2 components. BL₇ is depolarized by 20 mV to emphasize the hyperpolarizing phase seen in seawater. Addition of hexamethonium to the bath largely blocks the initial depolarizing phase, and d‐tubocurarine largely blocks the hyperpolarizing phase. The potential remaining following application of either drug is increased compared to the corresponding component recorded in seawater, presumably because the 2 underlying conductances overlap in time.