Chemically Mediated Transmission at a Giant Fiber Synapse in the

Document Sample
Chemically Mediated Transmission at a Giant Fiber Synapse in the Powered By Docstoc
					Published February 1, 1969




                             Chemically Mediated Transmission at
                             a Giant Fiber Synapse in the Central
                             Nervous System of a Vertebrate

                                   A. A. A U E R B A C H and M. V. L. B E N N E T T
                                   From the Department of Anatomy, Albert Einstein College of Medicine, Bronx, New York
                                   10461, and the Laboratory of Neurophysiology, Department of Neurology, College of
                                   Physicians and Surgeons, Columbia University, New York 10039. The present address of




                                                                                                                                    Downloaded from jgp.rupress.org on May 6, 2011
                                   both authors is Department of Anatomy, Albert Einstein College of Medicine, Bronx, New
                                   York 10461


                                   ABSTRACT The hatchetfish, Gasteropelecus, possesses large pectoral fin adductor
                                   muscles whose simultaneous contraction enables the fish to dart upwards at
                                   the approach of a predator. These muscles can be excited by either Mauthner
                                   fiber. In the medulla, each Mauthner fiber forms axo-axonic synapses on four
                                   "giant fibers," two on each side of the midline. Each pair of giant fibers in-
                                   nervates ipsilateral motoneurons controlling the pectoral fin adductor muscles.
                                   Mauthner fibers and giant fibers can be penetrated simultaneously by micro-
                                   electrodes close to the synapses between them. Electrophysiological evidence
                                   indicates that transmission from Mauthner to giant fiber is chemically medi-
                                   ated. Under some conditions miniature postsynaptic potentials (PSP's) are
                                   observed, suggesting quantal release of transmitter. However, relatively high
                                   frequency stimulation reduces PSP amplitude below that of the miniature po-
                                   tentials, but causes no complete failures of PSP's. Thus quantum size is reduced
                                   or postsynaptic membrane is desensitized. Ramp currents in Mauthner fibers
                                   that rise too slowly to initiate spikes can evoke responses in giant fibers that
                                   appear to be asynchronous PSP's. Probably both spikes and ramp currents
                                   act on the same secretory mechanism. A single Mauthner fiber spike is followed
                                   by prolonged depression of transmission; also PSP amplitude is little affected by
                                   current pulses that markedly alter presynaptic spike height. These findings
                                   suggest that even a small spike releases most of an immediately available store of
                                   transmitter. If so, the probability of release by a single spike is high for any
                                   quantum of transmitter within this store.


                                   INTRODUCTION
                             A n i m p o r t a n t p r o b l e m in the s t u d y of synapfic transmission is the r e l a t i o n
                             b e t w e e n pre- a n d postsynaptic potentials. A l t h o u g h this r e l a t i o n has b e e n
                             d e t e r m i n e d in a n u m b e r of instances of electrical transmission (5, 7, 8, 15,
                                                                                                                             ~83



                                                           The Journal of General Physiology
Published February 1, 1969




                                    I84             THE   JOURNAL   OF   GENERAL   PHYSIOLOGY   • VOLUME   53   "   1969

                             37), it is little known at most synapses where transmission is chemically
                             mediated, because intracellular electrodes cannot be placed in both pre- and
                             postsynaptic structures simultaneously. T h e only previously known exception
                             is the giant synapse of the squid (17, 23, 27, 28, 33) but considerable data
                             have also been obtained from the neuromuscular junction (21, 22), the chick
                             ciliary ganglion (29, 30), and certain electroreceptors (cf. reference 6).
                                This paper describes a chemically transmitting synapse between large
                             nerve fibers in the brain of the hatchetfish, Gasteropelecus. Although the fibers
                             .cannot be visualized in vivo, they can be penetrated by independently con-
                             trolled microelectrodes and identified using electrophysiological criteria.
                             Both pre- and postsynaptic structures can be recorded from simultaneously,
                             and the input-output relation resembles in most respects that demonstrated
                             or inferred for other chemically transmitting synapses. There are, however,




                                                                                                                           Downloaded from jgp.rupress.org on May 6, 2011
                             a n u m b e r of important differences, most of which can be explained by assum-
                             ing that a single presynaptic impulse releases a large fraction of an immedi-
                             ately available store of transmitter.
                                T h e hatchetfish, Gasteropelecus, is a c o m m o n aquarium fish imported from
                             South America, and is so named because of its characteristic shape. Its en-
                             larged, fan-shaped coracoid bones are fused and protrude to form attach-
                             m e n t sites for the powerful adductor muscles of the pectoral fins (Fig. 1 A,
                             Fig. 2). T h e fish is a surface feeder and its pectoral fins enable it to j u m p
                             appreciable distances into the air, in what is apparently a fast escape reac-
                             tion. It is also reported to use its pectoral fins to taxi along the surface with
                             only the ventral portion of the body submerged. It has even been said to fly
                             by flapping its fins rather than gliding as do other forms of flying fish (11,
                             38).
                               T h e chemically transmitting synapse described in this paper is involved in
                             the control of the pectoral fin adductor muscles. T h e paper following this one
                             describes the next lower synapse in the control system and the over-all reflex
                             activity. This second synapse is electrotonically transmitting and is the first
                             to be discovered in a vertebrate where the junctional m e m b r a n e rectifies.
                                Preliminary communications of some of this work have appeared (1, 2).
                                    M E T H O D S

                             Animals about 1~/~ inches in over-all length (probably G. sternicla) were employed
                             for most experiments. Curare (8-10 mg/kg) was given to prevent movement. The
                             medulla and upper spinal cord were exposed from the dorsal side. Respiration was
                             maintained by perfusion through the mouth with physiological saline for Electro-
                             phorus (25). The composition of this saline is 169 rnu NaC1, 5 mM KC1, 3 m_u CaCI~,
                             1.5 mu MgCI~, 1.2 mM Na2HPO~, and 0.3 mM NaH2PO4. Saline solution instead of
                             aquarium water was used for respiration because of difficulty in keeping the perfusate
                             out of the exposed region.
                                Conventional microelectrode techniques were employed. Intracellular stimula-
Published February 1, 1969




                             A. A. AUERBACHAND M. V. L. BENNETT Chgmically Transmitting Synapse            I8 5

                             tion was ordinarily carried out by means of a bridge circuit and] all illustrations of
                             intracellular stimulation except Fig. 3 F, Fig. 4 G-E, and Fig. 6 A, A' were obtained
                             using this technique. When the bridge was used, the membrane potential during the
                             applied current could be determined indirectly by measuring the change in ampli-
                             tude of a spike evoked during the current pulse (14). The relation between spike
                             height and current gives a measure of input resistance on the assumption that the
                             effective resistance at the peak of the spike is low compared to that at rest.
                                In most experiments, two independent microelectrodes were used. Usually a
                             grounded shield was inserted between them to reduce cross talk, and any remaining
                             contribution of cross talk due to spikes could be evaluated by grounding one or the
                             other electrode. When the two electrodes penetrated the same fiber, direct measure-
                             ments of potential during applied currents could be obtained. Electrotonic spread
                             was also measured by applying current through one electrode in a bridge circuit
                             while recording potential with a second electrode. In these experiments a spike was




                                                                                                                        Downloaded from jgp.rupress.org on May 6, 2011
                             evoked by spinal stimulation during the current pulse. As a function of the current,
                             Io, three voltages were measured: the change in spike amplitude at the polarizing
                             electrode, AVe° ; the change in spike amplitude at the second electrode, AVe~ ; and
                             the change in membrane potential at the second electrode, V,, the second electrode
                             being at a distance, x, from the first electrode. As seen below all three of these rela-
                             tions are linear over a sizeable range, and their slopes define resistances which can
                             be denoted as Reo, Re,, and R~, where for single measurements Reo = AVeo/Io, Rex =
                             AVe,/Io, and R , = V J I o . Rso and Re, are effective resistances and R, is a transfer
                             resistance (that is, the ratio of potential change at one point to inducing current
                             applied at a different point). T h e "true" input resistance, Ro, is the membrane po-
                             tential change recorded at the first electrode, Vo, divided by the current (Ro =
                             Vo/Io). R , and V, approach Ro and Vo as x becomes small (provided radial voltage
                             drops in the cytoplasm can be neglected), but where x is significant, Ro and Vo
                             can only be calculated from the directly measured resistances.
                                On the assumption that the change in spike amplitude is the same proportion of
                             the change in membrane potential at each electrode, one can write Vo as

                                                                  Vo -- A V . o V ~                               (1)
                                                                           AV°~

                             The same relation holds for the corresponding resistances as shown by dividing each
                             voltage of this equation by the value of polarizing current.

                                                                   Ro - R.oR,                                     (2)
                                                                            Rs~
                             This value of input resistance is corrected for the effective resistance at the peak of
                             the spike and for decrement due to separation of the two electrodes. T h e same data
                             allow calculation of a space constant, X, from the equation for electrotonic spread of
                             potential along a uniform core conductor:

                                                                  V~ =    Voe-~/x
Published February 1, 1969




                                     I86          THE   JOURNAL      OF    GENERAL   PHYSIOLOGY   • VOLUME   53   • I969


                             or


                                                                     =    x/In (Vo/Vx)                               (3)
                             Instead of using single pairs of voltage values, greater accuracy can probably be
                             obtained by using the resistances Rso and R ~ . From equation (1),

                                                                          Vo   Reo
                                                                          Vx   Rs~
                             thus,

                                                                  )~ = x/In (R,o/R,x)                                (4)
                                In some experiments, three independent microelectrodes were used. The caudal




                                                                                                                           Downloaded from jgp.rupress.org on May 6, 2011
                             spinal cord and right muscle nerve were stimulated using pairs of fine silver wire
                             electrodes 125/~ in diameter and insulated except at the tips. The electrodes were
                             placed close to the structure to be stimulated after making a small incision in the
                             skin.
                                     RESULTS

                                     Morphology

                             Hatchetfish possess two large M a u t h n e r fibers which run the length of the
                             spinal cord as in m a n y other species. At the level of the fourth ventricle, these
                             fibers are 40-60 /~ in diameter including the thick myelin sheath (Fig. 1
                             B-D). O n each side of the midline at this level there are also two other ex-
                             ceptionally large fibers which we have termed giant fibers. T h e cell bodies of
                             origin are located somewhat rostrally, but since the axons taper markedly in
                             this direction, the cell bodies have not yet been identified. Each giant fiber
                             forms several axo-axonic synapses with the ipsilateral M a u t h n e r fiber (Fig.
                              1 B - D ) usually by short (10-20 t~) myelinated processes from the M a u t h n e r
                             fiber, but there m a y also be a similar process from the giant fiber. Each
                             giant fiber has a large (about 30 # diameter) myelinated branch that crosses
                             the midline and passes dorsal to the contralateral M a u t h n e r fiber and then
                             terminates ventrolaterally in the neuropil (Fig. 1 B). Each transversely
                             running branch makes a single synaptic contact with a short process from
                             the M a u t h n e r fiber. T h e several ipsilateral synapses lie over an anterior-
                             posterior distance of less than 0.6 ram. T h e cross-branches run transversely
                             about 0.2-0.3 m m before synapsing on the contralateral M a u t h n e r fibers.
                             Electrophysiological evidence indicates that the space constants of both
                             M a u t h n e r and giant fibers are quite long. T h e morphological d a t a are in
                             agreement in that no naked axonal m e m b r a n e has been seen other than that
                             at the synapses, and the heavy myelin sheaths must provide substantial sur-
                             face insulation for the fibers.
Published February 1, 1969




                                            ~
                                        -~~.o
                                    :~o~                   .
                                    ~'~           ~.~
                                    0    ~        ~    ~
                                         ~" .~'~0


                                        ~         ~.~




                                    0    0    ~   TM




                                                               Downloaded from jgp.rupress.org on May 6, 2011
                                         o
                                    o    O~       "0   o

                                                  "0 "0
                                    0 ,-~ ,z:


                                    '~.~ -~
                                       ;:~ ~
                                                  "~   0




                                    0    ~         I   o




                                         0    m   m




                                        .-~ . ~



                                    ~~            ~

                                        ~o


                                        ~=~.~



                             I8 7
Published February 1, 1969




                                     i88                  THE JOURNAL OF GENERAL PHYSIOLOGY - VOLUME 53 " I969

                                From the synaptic region, the giant fibers send processes ventroposteriorly
                             t o t h e r e g i o n o f t h e l a r g e m o t o n e u r o n s t h a t i n n e r v a t e t h e p e c t o r a l fin a d -
                             ductor muscles. Direct contacts have not been seen between giant fibers and




                                                                                                                                                         Downloaded from jgp.rupress.org on May 6, 2011
                                     FmURE 2. Diagram of the fish and reconstruction of the relations between Mauthner
                                     fibers, giant fibers, and pectoral fin adductor motoneurons. The diagram of the fish
                                     shows the central nervous system, the pectoral fin adductor muscle (m), and its innerva-
                                     tion. The anterior of the muscle is supplied by a nerve running from the first spinal seg-
                                     ment (nl). Some caudal muscle fibers are innervated by a nerve from the second segment
                                     (n2). The coracoid bones underlie the entire muscle. In the medulla, each Mauthner
                                     fiber (m]) makes several synapses with each ipsilateral giant fiber (gf) and a single
                                     synapse with each contralateral giant fiber. The cross-branches of the giant fibers are
                                     paired; the pairs are about 100 t2 apart. Processes of each giant fiber synapse with each
                                     ipsilateral adductor motoneuron (mn) in the first spinal segment. There are about 40
                                     motoneurons on each side of the midline, but for clarity only 3 motoneurons on one side
                                     are shown. Transmission from Mauthner fiber to giant fiber is mediated by chemically
                                     transmitting synapses (cs). Electrically transmitting synapses (es) couple the giant fibers
                                     and ipsilateral motoneurons.

                             motoneurons but electrophysiological data establish a synaptic relationship
                             (3). T h e m o t o n e u r o n s i n n e r v a t i n g t h e a d d u c t o r m u s c l e lie p r i m a r i l y i n t h e
                             first s p i n a l s e g m e n t , a l t h o u g h t h e r e a r e a l s o a few in t h e s e c o n d s e g m e n t .
                             T h e a x o n s o f t h e m o t o n e u r o n s a r e a b o u t 20 ~ in d i a m e t e r w h i c h is l a r g e r
Published February 1, 1969




                             A. A. AUERBACHAND M. V. L. BENNETT ChemicallyTransmittingSynapse           I89

                             than the other axons in the peripheral nerve, and they are readily traced in
                             their course to the muscle (Fig. 2). T h e r e are 40-50 motoneurons on each
                             side as determined by counts of fibers in the ventral root. T h e M a u t h n e r
                             fibers, giant fibers, and several motoneurons on one side are d i a g r a m m e d in
                             Fig. 2.
                                  Identification and Properties of the Fibers
                             T h e M a u t h n e r fibers were penetrated in the medulla and first spinal seg-
                             ment, generally somewhat caudal to the region of the cross-branches of the
                             giant fibers. At this level, they were usually found about 50-100 ~ on either
                             side of the midline at a depth of 400-500 ~ from the surface. T h e resting
                             potential in the M a u t h n e r fibers was usually about 70 m y inside negative,
                             and the spike was about 80-90 m y in amplitude. T h e fibers were identified




                                                                                                                    Downloaded from jgp.rupress.org on May 6, 2011
                             by the short latency (about 0.3 reset) of their response to stimulation of the
                             caudal spinal cord (Fig. 3 A). This delay corresponded to a conduction
                             velocity of about 80 m/see, and simultaneous external recordings indicated
                             that these fibers were the most rapidly conducting and lowest threshold ele-
                             ments in the spinal cord. I n response to paired stimuli or brief tetani, they
                             could conduct impulses separated by as little as 1.2 msee (Fig. 3 B). T h e
                             rising phase of the spike was slightly faster than the falling phase, and neither
                             phase showed an inflection or "shoulder" in uninjured axons. T h e duration
                             at the base of the spike was about 0.5 reset. T h e spike was followed by a
                             brief hyperpolarizing afterpotential that often was separated from it by a
                             distinct inflection (arrow, Fig. 3 A). In two experiments, a M a u t h n e r fiber
                             identified by these characteristics was marked by intracellular iontophoretic
                             injection of methyl blue (34). Dissection following formalin fixation con-
                             firmed the electrophysiologieal identification.
                                 T h e giant fibers were also penetrated in the medulla and first spinal seg-
                             ment, usually at a depth somewhat greater than the M a u t h n e r fibers. T h e
                             resting potential was about 90 m y and the spike amplitude was often as
                             large as 120 Inv. Spikes evoked by stimulation of the caudal spinal cord had
                             a latency of about 0.7 rnsec (Fig. 3 C). In response to graded spinal stimuli,
                             a characteristic all-or-none component could be observed on the falling
                              phase of the response (arrow, Fig. 3 D). As shown below, this component
                              was the PSP due to activity of the higher threshold M a u t h n e r fiber (the
                              PSP produced by the lower threshold M a u t h n e r fiber having evoked a
                              spike at a lower stimulus strength). W h e n a pair of spinal stimuli was given
                              separated by an interval of about 5-100 msec the second spike in the giant
                              fiber failed, revealing the underlying PSP from the M a u t h n e r fiber (Fig. 3
                              E). This PSP could also be demonstrated by moderate hyperpolarization
                              (Fig. 3 F) or by repetitive stimulation at frequencies that often could be less
                              than 10/sec (Fig. 3 G). Another property was that graded depolarizations
Published February 1, 1969




                                       19o                THE       JOURNAL      OF    GENERAL        PHYSIOLOGY            • VOLUME       53   "   1969

                             c o u l d b e e v o k e d in a g i a n t fiber b y s t i m u l a t i o n o f t h e ipsilateral p e r i p h e r a l
                             n e r v e s (Fig. 3 H ) . T h e s e d e p o l a r i z a t i o n s w e r e a b o u t 0.4 m s e c in l a t e n c y ,
                             a n d o f t e n b e c a m e l a r g e e n o u g h to excite t h e g i a n t fiber• T h e y w e r e d u e t o




                                                                      •-                                    :g




                                   B                            D     !1                 F                           ~H




                                                                                                                                                             Downloaded from jgp.rupress.org on May 6, 2011
                                                                                             ,                   :   o        i:


                                                                                                                        J

                                                                                                              R

                                       Floum~. 3. Characteristic responses of Mauthner and giant fibers. A, the spike in a
                                       Mautlmer fiber evoked by spinal stimulation at the level of the dorsal fin. The arrow
                                       indicates a characteristic inflection preceding the afterhyperpolarizatlon. B, spikes in a
                                       Mauthner fiber evoked by a pair of spinal stimuli separated by 1.2 msec. C, a spike in
                                       a giant fiber evoked by spinal stimulation. D, superimposed sweeps showing an all-or-
                                       none component (arrow) on the falling phase of the spike in a giant fiber as the strength
                                       of spinal stimulation was varied. This component was the PSP due to excitation of the
                                       higher threshold Mauthner fiber. E, the responses in a giant fiber to paired spinal stimuli
                                       that excited both Mauthner fibers. Failure of the second spike occurred at an interval
                                       between stimuli of 10 msec. The middle portion of the sweep is omitted. F, upper trace,
                                       hyperpolarizing current applied in a giant fiber. Lower trace, potential recorded by a
                                       second microelectrode in the same fiber. The spike was evoked by spinal stimulation
                                       during the current pulses. Superimposed sweeps show the spike and the underlying
                                       PSP, which was revealed when the hyperpolarizing current was increased sufficiently
                                       to block the spike. G, superimposed sweeps showing failure of the spike in a giant fiber
                                       when the spinal cord was stimulated at a rate of about 10/sec. H, graded potentials in
                                       a giant fiber produced by graded antidromic stimulation of the nerve innervating the
                                       ipsilateral pectoral fin muscles. All time calibrations, 1 msec. Voltage calibrations in
                                       A-G, 50 my. In this and subsequent figures mf and gf signify voltages recorded from
                                       Mauthner fiber and giant fibers, respectively.


                             e l e c t r o t o n i c s p r e a d f r o m a n t i d r o m i c a l l y a c t i v a t e d m o t o n e u r o n s , as will b e
                             s h o w n in t h e f o l l o w i n g p a p e r (3). T h e f o r e g o i n g c h a r a c t e r i s t i c s w e r e u s e d to
                             i d e n t i f y a g i a n t fiber, a n d in t w o e x p e r i m e n t s , i o n t o p h o r e t i c i n j e c t i o n o f
                             m e t h y l b l u e a n d s u b s e q u e n t dissection c o n f i r m e d t h e i d e n t i f i c a t i o n .
Published February 1, 1969




                             A. A. At/Eli.BACHAND M. V. L. B~Nr~TT ChemicallyTransmittingSynapse        I9I

                                I n p u t resistances of M a u t h n e r and giant fibers were measured for hyper-
                             polarizing current applied through a single electrode in a bridge circuit (see
                             Methods). In six experiments, the input resistance of the M a u t h n e r fibers had
                             a m e a n value of 0.74 m e g o h m and ranged from 0.63 to 0.98 megohm. In five
                             experiments the m e a n input resistance of the giant fibers was 0.51 megohm,
                             with a range from 0.35 to 0.73 megohm. These input resistances as measured
                             would be somewhat lower than the actual values because of the finite re-
                             sistance at the peak of the spike.
                                A more accurate method using two intracellular electrodes permitted simul-
                             taneous measurement of both input resistance and space constant (see
                             Methods). In Fig. 4 B and 4 C are shown records from an experiment on a
                             M a u t h n e r fiber. T h e middle trace is the recording from one of the electrodes
                             in a bridge circuit, the upper trace is the potential recorded by the second




                                                                                                                     Downloaded from jgp.rupress.org on May 6, 2011
                             electrode 0.6 rnm away, and the lower trace is the polarizing current. A spike
                             was evoked during the current pulse by spinal stimulation. T h e graph of Fig.
                             4 A is from the same experiment and shows the change in spike amplitude
                             recorded by the polarizing electrode, AV 80, the change in spike amplitude at
                             the second electrode, AV,,, and the change in m e m b r a n e potential at the
                             second electrode, V,, all plotted against polarizing current, Io. These three
                             relations are linear. T h e calculated relation between the change in m e m b r a n e
                             potential at the polarizing electrode, Vo, and polarizing current (equation 2)
                             is the line in Fig. 4 A with the greatest slope; the slope of this line is the cor-
                             rected input resistance as discussed above. In this experiment the calculated
                             input resistance was 0.86 megohm. T h e space constant calculated from the
                             slopes of the change in spike amplitude at the two electrodes was 2.7 mm. In
                             two similar experiments the input resistances were 0.76 and 1.02 megohm_s
                             and the space constants were 3.6 and 2.9 mm. T h e means of the three values
                             of input resistance and space constant were 0.88 m e g o h m and 3.1 mm,
                             respectively.
                                 In three similar experiments on giant fibers, the input resistances were
                             0.38, 0.50, and 0.67 m e g o h m giving a m e a n value of 0.52 megohm. In the
                             same experiments, calculated values of the space constant were 3.2, 2.8, and
                             2.9 mm, respectively, giving a m e a n value of 3.0 mm. T h e values of input
                             resistance of the giant fiber are for hyperpolarizations less than about 30 mv
                              in which range the current voltage relations are linear. As shown in the
                              following paper, large hyperpolarizations cause a 20-30% increase in input
                              resistance of the giant fiber (V~/Io where x is small) and the current-voltage
                              relations become somewhat nonlinear.
                                 Fig. 4 D and 4 E show characteristic records obtained from the M a u t h n e r
                              fiber when linearly increasing or r a m p currents were applied through one
                              electrode while recording voltage through a second electrode. This type of
                              current application was used, as described below, to produce large depolariza-
Published February 1, 1969




                                 ~92                THE   JOURNAL    OF     GENERAL      PHYSIOLOGY          •   VOLUME     53      "    z969


                             A

                             xl~*4.         3       2     1     0

                             ompJ
                                                •               10
                                                                         , ,,            ,L            ~,.           ;
                                                                                                                                    |~
                                                                                                                                    I~
                                                                     D                             E                         ''~
                                   • v~,    /'/                 20

                                       V/Ro:O.86M~




                                                                                                                                                Downloaded from jgp.rupress.org on May 6, 2011
                                       v~                       3o
                                                              rnv

                                                                                      ' $ 0 msee                         5 0 msec
                              FIGURE 4. Input resistance and space constant of the Mauthner fiber. A-C, applica-
                              tion of rectangular current pulses. Sample records in B and {2, voltage-current relations
                              in A. Two electrodes separated by about 0.6 mm simultaneously penetrated the Mauth-
                              ner fiber. Rectangular hyperpolarizing current pulses were passed through one electrode
                              in a bridge circuit (current on lower traces, voltage on middle traces) while recording
                              directly with the second electrode (upper traces). Mauthner fiber spikes were evoked
                              during the current pulses (C) and when no current was passed (]3). Changes in spike
                              amplitude as a function of applied current, Io, were determined from recordings by the
                              second electrode (AV,,) and by the electrode in the bridge circuit (AV,o) and plotted in
                              A. Changes in membrane potential at the second electrode (Vx) were directly measured
                              and also plotted as a function of Io. The lines drawn through these three sets of points
                              were fitted by eye. Changes in membrane potential at the electrode in the bridge circuit
                              were recorded along with an unknown amount of potential due to bridge imbalance.
                              The actual change in membrane potential, Vo, shown in A, was calculated as described
                              in the methods (equation 1). (The values of Vo indicate that the bridge was fairly well-
                              balanced in C.) The calculated input resistance (equation 2) was Ro = 0.86 Mr2. The
                              calculated space constant (equation 4) was 2.7 ram. D-E, effects of linearly increasing
                              ramp currents. Separate experiments in D and E. Two electrodes separated by about
                              0.5 nun simultaneously penetrated the Mauthner fiber. R a m p currents (lower traces)
                              were applied through one electrode while recording voltage with the other electrode
                              (upper traces). D, two superimposed sweeps of equal amplitude hyperpolarizing and
                              depolarizing currents. The slopes of the hyper- and depolarizing potentials were equal
                              up until q- 30 my, corresponding to an input resistance of about 0.8 Mf]. At larger hy-
                              perpolarizing potentials, the hyperpolarizing resistance Vx/Io remained constant. At larger
                              depolarizing potentials, the depolarizing resistance Vx[I~, gradually decreased until at
                              about 60 mv it was down by about 30%. E, independence of depolarizing resistance of
                              the rate of current increase. Six superimposed sweeps. The voltage produced by a given
                              value of current was little affected when the rate of rise was changed over the fourfold
                              range illustrated. Calibrations the same for B and C.
Published February 1, 1969




                             A. A. AUE~ACHAND M. V. L. BENNETT Chgmictgll7 TransmittingSynapse       193

                             tions in the Mauthner fiber without exciting a spike. The input resistances for
                             currents of both polarities were approximately equal over the first 30 my of
                             potential change (Fig. 4 D). When hyperpolarizing currents were further
                             increased, the resistance remained constant; i.e., the voltage trace remained
                             linear with the same slope. When depolarizing currents were further increased,
                             the resistance began to decrease; i.e., the slope of the voltage trace decreased.
                             T h e decrease in depolarizing resistance (V,/Io) was presumably due to de-
                             layed rectification, and averaged about 20-30% when the potential was about
                             60-100 mv above the resting potential (Fig. 4 D and E). T h e rate of change of
                             polarizing current had little effect on the shape of the voltage trace (Fig. 4 E)
                             except when ramps were rising sufficiently rapidly to initiate spikes.

                                  Relation between Mauthner and Giant Fibers




                                                                                                                 Downloaded from jgp.rupress.org on May 6, 2011
                             The synaptic relation between Mauthner and giant fibers was established in
                             experiments in which the fibers were simultaneously penetrated. T h e Mauth-
                             ner fiber was presynaptic, i.e. a directly evoked spike in the Mauthner fiber
                             produced a PSP in the giant fiber (Fig. 5 A), whereas a directly evoked spike
                             in a giant fiber did not lead to a PSP in the Mauthner fiber (Fig. 6 B). T h e
                             PSP usually initiated a spike (Fig. 5 A and B), but hyperpolarizing the giant
                             fiber (Fig. 5 E) or repetitively stimulating the Mauthner fiber (Fig. 7) could
                             cause failure of impulse initiation and demonstrate the underlying PSP. T h e
                             latency of the PSP was 0.3-0.4 msec measured from onset of the directly
                             evoked spike in the Mauthner fiber to onset of the PSP in the giant fiber
                             (arrows, Fig. 5 A).
                                As indicated in Fig. 2, a Mauthner fiber activates both ipsilateral and
                             contralateral giant fibers. This relation was shown in many experiments using
                             two electrodes where it could be clearly seen on which side of the midline the
                             penetrated ,Mauthner and giant fibers lay. A directly evoked spike in a
                             Mauthner fiber was always followed by a PSP in a giant fiber whether the
                             fibers were ipsi- or contralateral. This relation between fibers was further
                             demonstrated using three electrodes. In two experiments, a Mauthner fiber
                             and one ipsilateral and one contralateral giant fiber were recorded from
                             simultaneously. Direct stimulation of the Mauthner fiber excited both giant
                             fibers (Fig. 5 C). In two additional experiments, both Mauthner fibers were
                             penetrated while simultaneously recording in a giant fiber. Direct stimulation
                             of each Mauthner fiber produced a PSP in the giant fiber which was of
                             necessity ipsilateral to one Mauthner fiber and contralateral to the other
                             (Fig. 5 F).
                                Stimuli were also applied to the spinal cord while simultaneously recording
                             from Mauthner and giant fibers (Fig. 5 B, D, G, and H). As the stimulus
                             strength was increased, excitation of a Mauthner fiber was always followed by
                             a corresponding PSP component in the giant fiber. When the PSP from the
Published February 1, 1969




                                 I94                THE    JOURNAL     OF        GENERAL        PHYSIOLOOY          • VOLUME     53       '   z969

                             A         ~             C mr---                E                    /              G      .~

                                                :                                      o,.,t .\. i                     !i.o,
                                 _+                                     ,+                                      _                     I,,
                                              "+"~' " - " " - - "     "-"+        '+       . . . .                  ~ +        "......1
                             B                       D                       F                              H         !l
                                                                                       ,                              ~l"gf
                                                                                 -+.                                m,-                 I+
                                                               ~'++     '+ ~'/W--~----,,                   I~        ~ju+          ,,




                                                                                                                                                     Downloaded from jgp.rupress.org on May 6, 2011
                                                         ._..+~        ,,         _                         :
                                                                                                     + Imsec.g .~\              ..... IP;
                                            lmsec'                    ms"~C
                                                                      1

                                 FIGURE 5. Synaptic relation between Mauthner and giant fibers. A, a directly evoked
                                 spike in a Mauthner fiber (upper trace, current on the lower trace) was followed by a
                                 spike in a giant fiber (middle trace); i.e., the Mauthner fiber was presynaptic. B, same
                                 fibers and display as in A. Spinal stimulation evoked spikes in the Mauthner and giant
                                 fibers that were separated by the same interval as when the Mauthner fiber was directly
                                 excited. C, a directly evoked spike in a Mauthner fiber (first trace, current on the fourth
                                 trace) evoked spikes in both an ipsilateral and a contralateral giant fiber (second and
                                 third traces). D, same fibers and display as in C, but spikes were evoked by spinal stim-
                                 ulation. Activation of the giant fibers by both Mauthner fibers is indicated by the occur-
                                 rence of an additional component on the falling phase of each giant fiber spike. E, a
                                 directly excited spike in a Mauthner fiber (upper trace) produced a PSP in a giant
                                 fiber (middle trace) that initiated the giant fiber spike. The PSP was demonstrated by
                                 hyperpolarizing the giant fiber (current on lower trace, two superimposed sweeps in
                                 one of which the giant fiber spike was blocked). F, PSP's produced in a giant fiber by
                                 direct excitation of each Mauthner fiber. The two upper traces show directly evoked
                                 spikes in each Mauthner fiber; the third trace shows the PSP's in the giant fiber, and
                                 the bottom trace shows the intracellular current applied to one Mauthner fiber. G,
                                 upper and lower traces, recording from each Mauthner fiber; middle trace, recording
                                 from a giant fiber. Spinal stimulation. The threshold of the Mauthner fiber recorded
                                 on the lower trace was somewhat lower than that of the other Mauthner fiber. At thresh-
                                 old for this fiber, the giant fiber was excited only when the Mauthner fiber was excited
                                 (superimposed sweeps showing the Mauthner fiber excited and not excited). The in-
                                 flection on the falling phase of the Mauthner fiber spike is an artifact caused by capaci-
                                 tative coupling between the microelectrodes which had not been adequately shielded.
                                 H, same fibers and display as in G, but stimulation at threshold for the other Mauthner
                                 fiber. When the second Mauthner fiber was excited, an additional component appeared
                                 on the falling phase of the giant fiber spike. All time calibrations 1 msec. All voltage
                                 calibrations 50 mv unless otherwise indicated.
Published February 1, 1969




                                            AND M. V. L. BENNETT ChemicallyTransmittingSynapse
                             A. A. AUERBACI-I                                                                I95

                             lower threshold M a u t h n e r fiber initiated a spike (Fig. 5 G), the second M a u t h -
                             ner fiber produced an additional component on the falling phase (arrow, Fig.
                             5~H; cf. Fig. 3 D, Fig. 5 G and D).
                                T h e latency of PSP's following spinal stimulation was identical to that for
                             direct stimulation of the M a u t h n e r fibers, provided a correction was m a d e for
                             the different time course in reaching the threshold of the M a u t h n e r fiber spike
                             (arrows, Fig. 5 A and B). Although there are several ipsilateral synapses and
                             only one contralateral synapse, the PSP components from the two M a u t h n e r
                             fibers were usually of about the same size. Occasionally, the amplitudes could
                             differ by a factor of up to four, but it was not determined which M a u t h n e r
                             fiber produced the smaller PSP's.

                                  Mode of Transmission




                                                                                                                          Downloaded from jgp.rupress.org on May 6, 2011
                             A n u m b e r of observations indicate that transmission at the M a u t h n e r fiber,
                             giant fiber synapse is chemically mediated. T h e most important evidence can
                             be summarized as follows: (a) T h e PSP could be inverted by sufficiently large
                             outward (depolarizing) currents. (b) No electrotonic coupling could be meas-
                             ured between the M a u t h n e r and giant fibers. (c) There was a delay of about
                             0.4 msec between the presynaptic spike in the M a u t h n e r fiber and the PSP
                             in the giant fiber. (d) At low to moderate frequencies of stimulation, the
                             presynaptic spike remained constant in amplitude, but the PSP could vary
                             randomly. O n the other hand, controlled variation of the amplitude of the
                             presynaptic spike produced little or no change in the average amplitude of
                             the PSP. (e) In certain circumstances, there were indications of transmitter
                             release in discrete packets or quanta, as has been observed at a n u m b e r of
                             chemically transmitting synapses (18, 29). These observations are discussed
                             more fully below.
                                Inversion of the PSP by outward currents is illustrated in Fig. 6 A. For these
                             experiments it was necessary to pass currents too large to allow the use of the
                             bridge circuit. Therefore, separate recording and current-passing electrodes
                             were placed in the giant fiber and the PSP was evoked by spinal stimulation.
                             T h e inversion indicates that there is a conductance increase associated with
                             generation of the PSP that is more or less independent of the potential across
                             the m e m b r a n e (4, 16). An electrically mediated PSP cannot show this in-
                             version, and the inference is that transmission must be chemically mediated.
                             Measurement of the reversal potential of the PSP was complicated by the
                             increased conductance due to the polarizing current. In Fig. 6 A, no PSP was
                             observed when the potential was about 90 m y positive to the resting potential.
                             In two other experiments, the measured reversal potential was about 100 mv
                             above the resting potential. F r o m these experiments and the estimated resting
                             potentials, the reversal potential was probably close to zero m e m b r a n e
                             potential.
Published February 1, 1969




                                                                          B



                                                                              -- i,m,               [!


                                                          2 msec
                                                                   ii                               o

                                                 .i
                                    C

                                                                                  !!--mr       ;    I~




                                                                                                                         Downloaded from jgp.rupress.org on May 6, 2011
                                                                                           2 msec

                             FIo~J~ 6. Reversal of PSP's in the giant fiber and absence of electrotonic coupling
                             between Mauthner and giant fibers. A, inversion of the PSP by depolarizing current
                              (five traces superimposed photographically). A spike was evoked in the giant fiber by
                             spinal stimulation. Directly excited spikes can be distinguished at the beginning of the
                             two weakest depolarizing pulses and the amplitude of the subsequent spikes diminished
                             in these records. The PSP was not detectable about 90 mv above the resting potential
                             which presumably was close to the reversal potential. Inverted PSP's are seen on the
                             two upper traces. A', augmentation of the PSP by hyperpolarizing current (six super-
                             imposed traces, one without polarizing current). The orthodromic spike was blocked
                             by the smallest current applied, and increasing hyperpolarization increased the ampli-
                             tude of the PSP. B-C, the absence of electrotonic coupling between Mauthner and giant
                             fibers. B, a spike in a giant fiber (middle trace) was directly evoked by depolarizing
                             current (lower trace), but produced no measureable potential in the Mauthner fiber
                             (upper trace). C, approximately 0.1 /~amp of hyperpolarizing current in the Mauthner
                             fiber (upper trace) blocked propagation of a Mauthner fiber spike evoked by spinal
                             stimulation (bottom trace). Based on resistance measurements from other fibers this
                             current would have produced about 60-100 mv of hyperpolarization. No measurable
                             hyperpolarization was recorded in the giant fiber (middle trace). The spike in the giant
                             fiber was unaffected. Presumably, the other Mauthner fiber produced the PSP that
                             initiated the giant fiber response since with the usual sites of electrode placement, and
                             ff only a single Mauthner fiber were active, block of a spinally evoked spike in the
                             Mauthner fiber blocked the PSP in a giant fiber (Fig. 10 D). D, approximately 0.1 #amp
                             of hyperpolarizing current (upper trace) blocked the giant fiber spike evoked by spinal
                             stimulation (bottom trace). This current would have produced from 40-80 my hyper-
                             polarization. There was no measurable hyperpolarization in the Mauthner fiber (middle
                             trace) and the Mauthner fiber spike was unaffected. Calibrations the same in C and D.
                                                                    x96
Published February 1, 1969




                             A. A. AUER~ACHAND M. V. L. BENN~Ta" Chemically Transmitting Synapse                I97

                                Hyperpolarization augmented the PSP (Fig. 6 AP), as is observed at m a n y
                             chemically transmitting synapses, but as discussed in the following paper (3),
                             this property can also be exhibited at an electrotonic synapse. An estimate of
                             the PSP reversal potential can be obtained from extrapolation of the change
                             in PSP amplitude as a function of hyperpolarization. In Fig. 6 A' this value
                             is only a few millivolts positive to the resting potential and in other experi-
                             ments, the values ranged between 30 and 50 mv positive to the resting po-
                             tential. Two factors probably contributed to the discrepancy between these
                             estimated values and the directly measured ones. First, the measured values
                             m a y be somewhat high because delayed rectification decreased the space
                             constant, and the electrodes were at some distance from the synapses. Second,
                             there was some increase in input resistance of the giant fiber as it was hyper-
                             polarized (3). This change would have increased the degree of augmentation




                                                                                                                             Downloaded from jgp.rupress.org on May 6, 2011
                             of the PSP produced by hyperpolarization, and caused the extrapolated
                             reversal potential to be too low.
                                T h e degree of electrotonic coupling m a y be described in terms of the
                             coupling coefficients (the ratio of voltage in the second cell to voltage in the
                             first cell when current is applied in the first cell). T h e coupling coefficients
                             for hyperpolarization were always less than the measurable limit of about
                             0.005 whether current was applied in M a u t h n e r or giant fibers (Fig. 6 C and
                             D). Depolarization that evoked spikes in a giant fiber caused no depolarization
                             in a M a u t h n e r fiber (Fig. 6 B), but, of course, a spike in a M a u t h n e r fiber was
                             followed by a PSP in a giant fiber.
                                T o be valid, the measurement of synaptic delay and the demonstration of
                             absence of coupling require that the electrodes be close to the synaptic region.
                             In these experiments, the electrodes were always less than 0.5 m m apart and
                             close to the synapses as judged by both their proximity to the fourth ventricle
                             and the presence of a large PSP in the giant fiber. As noted above, the cal-
                             culated space constants in both M a u t h n e r and giant fibers are about 3 m m
                             and the conduction velocity in the M a u t h n e r fiber is very high. T h e processes
                             forming the actual synapses are smaller in diameter than the main parts of
                             the fibers, but calculations given in the discussion indicate that they are too
                             short to allow for significant decrement or conduction time.
                                T h e presence of synaptic delay is characteristic of chemically mediated
                             transmission although comparable delays can occur in electrotonic transmis-
                             sion (5). T h e delay measured at chemically transmitting synapses in cold-
                             blooded forms is usually 0.5 msec or greater at 20-25°C (cf. reference 8). At
                             the frog neuromuscular junction there is a m i n i m u m delay at 20°C of 0.4
                             msec measured from the peak negativity of the externally recorded presynaptic
                             spike to the onset of the PSP (19, 20). T h e value of the synaptic delay in
                             hatchetfish would be slightly shorter if measured in the same way.
Published February 1, 1969




                                       I98                    THE    JOURNAL          OP   GENERAL         PHYSIOLOGY                  - VOLUME   53   " I969


                                      Effects o/ Repetitive Stimulation
                              A s n o t e d a b o v e , t r a n s m i s s i o n a t t h e M a u t h n e r fiber, g i a n t fiber s y n a p s e w a s
                              f a t i g u e d b y r e p e t i t i v e s t i m u l a t i o n a t r a t h e r l o w f r e q u e n c i e s (Fig. 3 G ) . T h i s

                                                               Ai                       A2                A3        !
                                                                    --:.---      io            so
                                                                                               ---             1    ':            ac




                                                                                                                         5 m$oc




                                                                                                                                                                  Downloaded from jgp.rupress.org on May 6, 2011
                                                                       I         I         I         t         I            I
                                                                      0         10        20       30          40          C:O
                                                                                        STIMULUS NUMBER

                                      Fmum~ 7. Effects of repetitive stimulation on PSP amplitude. A, postactivation de-
                                      pression following a single directly evoked spike in a Mauthner fiber. Upper trace,
                                      depolarizing current in the Mauthner fiber; middle trace, recording in the Mauthner
                                      fiber; lower trace, recording in a giant fiber. Pairs of stimuli separated by varying in-
                                      tervals were given at about one pair per sec. The two stimuli and the evoked spikes in
                                      the Mauthner fiber remained of constant amplitude in all records. Ax, the PSP due to
                                      the first spike in the Mauthner fiber initiated a spike in the giant fiber and this part of
                                      the sweep is omitted in subsequent records. The PSP due to the second stimulus 10
                                      msec later was only about 4 my in amplitude. A2, when the interval between stimuli
                                     was about 50 msec, the second PSP recovered to about 6 mv. As, when the interval be-
                                     tween stimuli was about 100 msec, the second PSP became threshold for a giant fiber spike
                                     (two superimposed sweeps with and without a spike). Recovery was incomplete because
                                     the PSP initiating the spike in Ax rose faster than the PSP in As. Calibrations in Am B,
                                     PSP amplitude in a giant fiber during a train of 51 directly excited Mauthner fiber
                                     spikes separated by intervals of 160 msec. The PSP remained below threshold for the
                                     giant fiber (about 15 my) after the initial response. The three PSP's in response to the
                                     second through fourth stimuli successively increased in amplitude. The next 10 responses
                                     varied widely in amplitude, perhaps periodically. The remaining responses appear to
                                     have varied randomly except for a small downward trend.
                             r e d u c t i o n in t h e P S P ( p o s t a c t i v a t i o n d e p r e s s i o n ) w a s s t u d i e d b y g i v i n g p a i r s
                             o f s t i m u l i w i t h p e r i o d s o f 1 - 5 sec b e t w e e n pairs. P S P ' s d u e to t h e s e c o n d o f a
                             p a i r o f d i r e c t l y e v o k e d M a u t h n e r fiber spikes failed to excite a g i a n t fiber spike
                             a t i n t e r v a l s b e t w e e n s t i m u l i as l a r g e as 1 0 0 - 5 0 0 msec. W h e n t h e s e c o n d
                             M a u t h n e r f i b e r spike f o l l o w e d t h e first a t successively s h o r t e r intervals, t h e
Published February 1, 1969




                             A. A. AUERBACHAND M. V. L. BENNETT ChemicallyTransmitting Synapse              199

                             P S P decreased in amplitude (Fig. 7 A). If pairs of spinal stimuli were used
                             that were strong enough to excite both M a u t h n e r fibers, the reduction of the
                             second PSP to the same amplitudes as observed with direct stimulation of a
                             single M a u t h n e r fiber required shorter intervals between stimuli, since PSP's
                             from the two M a u t h n e r fibers summated. W h e n a pair of spinal stimuli was
                             separated by the shortest interval that still permitted excitation of both
                             M a u t h n e r fibers, the second PSP was only a b o u t 2-3 m v in amplitude. Except
                             for the first few milliseconds of the period of postactivafion depression, there




                                             mf--




                                                                                                                         Downloaded from jgp.rupress.org on May 6, 2011
                                             gfm~/       ~                        m    ~




                                             mf--               "1"                         '

                                                                                                I m,~¢

                                  Fmug~ 8. Decrease in PSP amplitude with increasing frequency of stimulation.
                                  Upper traces, Mauthner fiber spikes excited by spinal stimulation; lower traces, re-
                                  sponses in a giant fiber. The stimulation frequency was continuously increased from
                                  about 10/sec (A) to about 40/sec (F). PSP amplitude decreased steadily to a value of
                                  about 0.15 my. The PSP was always detectable, and there were no complete failures of
                                  transmission. Calibrations in F.

                             was no alteration in the M a u t h n e r fiber spike. T h e reduction in the P S P
                             cannot be attributed to increased conductance of the postsy-naptic cell, since
                             excitability of the giant fiber measured b y direct stimulation was unchanged
                             following the PSP (other than for a brief period of refractoriness if the first
                             P S P initiated a spike). Furthermore, stimulation of one M a u t h n e r fiber had
                             no effect on the P S P produced b y stimulation of the other provided a spike
                             was not evoked in the giant fiber. A possible explanation of the depression of
                             transmission seen with paired stimulation is that the transmitter immediately
                             available for secretion is depleted b y the first stimulus. Recovery of PSP
                             amplitude following such depletion would then be a consequence of replenish-
                             ment of this transmitter.
                                T h e effects of previous activity on transmission were also studied by giving
                             trains of stimuli at various frequencies separated by periods of rest. At moder-
Published February 1, 1969




                                   200             THE   JOURNAL    OF       OENERAL     PHYSIOLOOY       •    VOLUME   53   "   x969

                             ate frequencies of stimulation, the second P S P was greatly reduced (as in Fig.
                             7 A) b u t the amplitude of subsequent PSP's recovered to some extent (Fig.
                             7 B). In the next 10 or so responses, the PSP amplitude varied quite widely.
                             These variations m a y have had a periodic component with a frequency of
                             2-4/sec. After several seconds, the variability decreased b u t in the steady state
                             there continued to be considerable variation. Although further study is
                             required, the amplitude histogram is probably unimodal and symmetrical

                                              AI                   B!                    C t




                                               5X--,




                                                                                                                                        Downloaded from jgp.rupress.org on May 6, 2011
                                                  :                 i , ,H                 J




                                               ....2          ~. ,.i               ,__   _2.           , - ,o~
                                                                                                      5rnseC

                                  FmURE 9. Effect of prolonged stimulation on the PSP. Upper trace, recording in a
                                  Mauthner fiber; middle trace, recording in a giant fiber; lower trace, stimulating cur-
                                  rent. Following prolonged repetitive stimulation of the Mauthner fiber at 10-20/sec,
                                  the rate of stimulation was reduced to about 1/see. The PSP's in the giant fiber were
                                  broadened, revealing small components similar to "quanta" seen at other chemically
                                  transmitting synapses. Calibrations in F.

                             about the mean. It does not appear to fit a Poisson distribution, because the
                             dispersion of amplitudes is much smaller than would be expected for this
                             distribution.
                                T h e mean amplitude of the P S P during the steady state varied inversely
                             with frequency (Fig. 8). After several minutes of stimulation at frequencies of
                             10-20/sec, the PSP often broadened (Fig. 9) revealing m a n y small compo-
                             nents resembling " q u a n t a " seen at a number of chemically transmitting
                             synapses (9, 18, 26, 29, 31). After a short period of low frequency stimulation
                             or rest, the shape of the PSP returned to normal. T h e changes observed sug-
                             gest a desynchronization in the release of transmitter. W h e n resolvable, the
                             components were about 0.3-0.5 mv in amplitude. As described below, similar
                             components could be evoked by depolarization of the M a u t h n e r fiber. U n d e r
                             normal conditions, the rate of spontaneous occurrence of small potentials
Published February 1, 1969




                             A. A. AUERBACHAND M. V. L. BEm~m'l'r ChemicallyTransmittingSynapse          2ox

                             resembling the PSP components in Fig. 9 was no more than a few per second.
                             During a period of desynchronized release, these small potentials occurred
                             m u c h more frequently, and one can be seen at the start of the sweep in Fig.
                             9 D. Probably these potentials were due to spontaneous release of transmitter,
                             but the possibility that they were PSP's produced by impulse activity in other
                             neurons was not excluded.
                                W h e n the frequency of spinal or direct stimulation of the M a u t h n e r fiber
                             was increased to 20-40/sec, PSP's in the giant fiber further diminished in
                             amplitude (Fig. 8 A-F). T h e PSP's exhibited a continuous range of amplitudes
                             and after some seconds of stimulation approached the amplifier noise level,
                             which was usually about 50-100 #v. No "failures" of transmission were ob-
                             served; that is, there was always at least a small PSP. T h e coefficient of
                             variation (standard deviation divided by the mean) did not increase, as it




                                                                                                                      Downloaded from jgp.rupress.org on May 6, 2011
                             would have been expected to do if the reduction in PSP size were due to
                             reduction in the n u m b e r of quanta released. Assuming that the 0.3-0.5 mv
                             components observed correspond to normal sized quanta of transmitter, either
                             the size of a q u a n t u m or its postsynaptic action must be considerably reduced
                             at higher frequencies of stimulation.

                                  Relationship between Presynaptic Potential and PSP Amplitude
                             T h e PSP in a giant fiber was unaffected by current pulses that altered the
                             amplitude of a M a u t h n e r fiber spike over a wide range. T h e amplitude of a
                             propagated spike in a M a u t h n e r fiber could be changed -4-25% by current
                             pulses applied close to the synaptic region, but these changes had no effect on
                             the amplitude or time course of the PSP's. T h e amplitude variations normally
                             observed (Fig. 8) were still present, but their m e a n and distribution were
                             changed little, if at all, as m a y be seen from the superimposed sweeps in Fig.
                             10 A-C. In this same experiment, 50 measurements of PSP amplitude were
                             also m a d e for each case, i.e. with no current applied in the M a u t h n e r fiber,
                             as in Fig. 10 A; with depolarizing current, as in Fig. 10 B; and with hyper-
                             polarizing current, as in Fig. 10 C. T h e means and amplitude distributions in
                             each case were essentially identical. Fig. 10 E and 10 F from another experi-
                             m e n t show the absence of an effect in successive sweeps with and without
                             polarization. These records are representative of m a n y additional trials. In
                             these experiments, propagation in one M a u t h n e r fiber was blocked by injuring
                             it with a coarse microelectrode, and then the other M a u t h n e r fiber was
                             penetrated. T h e spinal cord was stimulated at a rate adequate to cause failure
                             of impulse initiation in the giant fiber. T h e recorded PSP resulted solely from
                             activity of the penetrated M a u t h n e r fiber, because if propagation in it was
                             blocked by hyperpolarization, the PSP failed completely (Fig. 10 D). T h e
                             range over which spike height could be varied was limited to the changes
                             produced by subthreshold depolarization and by hyperpolarization insuffi-
Published February 1, 1969




                             202             THE    JOURNAL     OF   GENERAL   PHYSIOLOGY          • VOLUME        5~   "   z969


                             A                        C                             E
                                                                                              ii         I

                                                                                                              f




                                                                                          mfm




                             B                        D




                                                                                                                                   Downloaded from jgp.rupress.org on May 6, 2011
                                                          "it                            '                    '"
                                     I          i               P~"~               o                                    J
                                                          k


                                                                                                           lmsec
                             FmuR~ 10. The lack of dependence of PSP amplitude on presynaptic spike height. A
                             single Mauthner fiber was activated by spinal stimulation (see text) at a rate sulticienfly
                             fast to reduce the PSP below threshold for excitation of the giant fiber. Stimulation was
                             maintained at this rate long enough for mean PSP amplitude to reach a steady-state
                             value. The amplitude of the spike in the Mauthner fiber was augmented by hyper-
                             polarizing currents (C and E), or diminished by depolarizing currents (B and F). Po-
                             larizing currents are shown on the upper trace in each case. The PSP's in the giant
                             fibers were recorded on the middle traces in A-D but during the pulses the lower traces
                             recording the Mauthner fiber potentials crossed over the middle traces. The middle and
                             lower traces in E and F are recordings from the Mauthner and giant fibers, respectively.
                             There are superimposed sweeps in A-C. A, the PSP due to the normal Mauthner fiber
                             spike showed random variations. B, depolarizing current decreased the presynaptic
                             spike height by about 25%. The mean amplitude and variability of the PSP were ap-
                             parently unchanged. C. hyperpolarizing current increased the presynaptic spike ampli-
                             tude by almost 30%. The mean amplitude and variability of the PSP showed little
                             change. D, demonstration that the PSP was due solely to the spike in the one Mauthner
                             fiber. A spinal stimulus adequate to excite both Mautlmer fibers was given. A strong
                             hyperpolarization caused a large component of the recorded spike to fail indicating that
                             propagation along the fiber was blocked. Correspondingly, the PSP in the giant fiber
                             failed completely. E and F, successive sweeps superimposed with and without polarizing
                             currents. E, hyperpolarizing current slightly delayed the presynaptic spike and in-
                             creased its amplitude about 20%. The PSP in the giant fiber was slightly delayed, but
                             it was unchanged in amplitude. F, depolarizing current diminished the amplitude of
                             the presynaptic spike by about 20%. The PSP was unchanged. Calibrations identical
                             within each series.
Published February 1, 1969




                             A. A..&trERBACHAND M. V. L. BENN~.TT ChemicallyTransmittingSynapse           003

                             cient to block propagation. T h e amplitude variation of the spike at the synapse
                             must have been approximately equal to that recorded. As already noted in
                             respect to the absence of electrotonic coupling, there would have been little
                             decrement of the hyperpolarizing potentials in reaching the terminals. T h e
                             depolarizations in these experiments would also have shown little decrement,
                             because they must have been too small to cause appreciable delayed rec-
                             tification (cf. Fig. 4 D and E).
                                T h e relative independence of PSP amplitude from presynaptic spike height
                             contrasts markedly with results of similar experiments on the squid synapse
                             in which the PSP was greatly affected (33). T h e results in the hatchetfish sug-
                             gest that maximal secretion of transmitter is evoked by even a small spike;
                             i.e., the secretory processes are easily saturated. One mechanism that would
                             explain the data is that even a small spike causes sufficient secretion to deplete




                                                                                                                       Downloaded from jgp.rupress.org on May 6, 2011
                             the immediately available transmitter store and that mobilization of addi-
                             tional transmitter is relatively slow. This explanation is also consistent with
                             the observation that a single stimulus produces a pronounced depression of
                             transmission.
                                T h e relation between pre- and postsynaptic potentials was also studied by
                             using linearly rising (ramp) depolarizing currents which increased at a rate
                             too slow to initiate spikes. This procedure allowed controlled depolarization of
                             the M a u t h n e r fiber in the range between a subthreshold pulse and a m i n i m u m
                             sized spike. These experiments were carried out using one electrode in the
                             M a u t h n e r fiber and one in the giant fiber. T h e currents required changed the
                             electrical properties of the polarizing electrode in the M a u t h n e r fiber and
                             during the ramps prevented accurate potential recording from this electrode
                             by means of the bridge circuit. Nonetheless, when a too rapidly rising r a m p
                             initiated a spike, the response could be seen on the bridge recording and was
                             also signalled by the occurrence of a large PSP in the giant fiber. As shown in
                             Fig. 4 D and E a r a m p current produced a more or less proportional r a m p
                             potential change.
                                W h e n sufficiently large r a m p depolarizing currents were applied in a
                             M a u t h n e r fiber, these stimuli evoked depolarizations in a giant fiber that
                             presumably resulted from the release of transmitter (Fig. 11). T h e responses
                             appeared to consist of m a n y small components (Fig. 11 F) like those seen
                             following prolonged repetitive stimulation (Fig. 9). T h e responses were
                             graded, and larger stimuli evoked larger responses (Fig. 11 A1-A4) until a
                             m a x i m u m was reached (Fig. 11 A4 and C). T h e responses began at approxi-
                             mately the same current value when the slope of the r a m p was changed (Fig.
                             11 BrB4). However, the more rapidly rising ramps evoked more synchronous
                             responses, which is consistent with the graded increase in response amplitude
                             as the r a m p current was increased. By comparison with experiments in which
                             two electrodes in a M a u t h n e r fiber were used to determine the potentials
Published February 1, 1969




                                     2o4                  THE       JOURNAL     OF   GENERAL           PHYSIOLOGY           • VOLUME         53   "   ~969

                             p r o d u c e d b y r a m p c u r r e n t s , t h e v o l t a g e t h r e s h o l d for      secretion was about
                             2 5 - 3 0 m v a b o v e t h e r e s t i n g p o t e n t i a l (Fig. 11 A2). T h i s          v a l u e is close t o t h a t
                             o b s e r v e d a t t h e s q u i d s y n a p s e (23, 97). W h e n a r a m p               c u r r e n t c o n t i n u e d to
                             rise b e y o n d t h e t h r e s h o l d , t h e responses c o n t i n u e d for a          period of 15-30 msec




                                                                                       B3


                                  °Jl                           ./%___




                                                                                                                                                              Downloaded from jgp.rupress.org on May 6, 2011
                                     C         ~                D                       E                            F        .,.~, i ~

                                                          _                                        .    .   .   .




                                                                                                                20msec                    2msee
                                     I~GURE 11. The effects of ramp currents applied in the presynaptic fiber. The stimuli
                                     (lower traces) were applied in the Mauthner fiber at a rate of repetition of 0.20/sec.
                                     The slope was kept below that required to initiate spikes. Responses in the giant fiber
                                     (upper traces) appeared to consist of bursts of small potentials which summated and
                                     reached a maximum amplitude of 4-5 my. All data in this figure were taken from the same
                                     fibers. A1-A4, ramp currents of increasing amplitude and constant slope. Small responses
                                     were evoked when the current reached a value of about 30 namp (A2). Larger and more
                                     synchronous responses were evoked by larger currents (As and A4). B1-B4, ramp currents
                                     of decreasing slope and constant peak amplitude. The slope was decreased by a factor
                                     of about 5 from B1 to B4. The responses were more dispersed in time when the ramp was
                                     more slowly rising. However, the responses began at approximately the same current
                                     level in each case. C, D, the responses became very small after about the same period
                                     whether the ramp current continued to rise (C) or whether it was held at a plateau value
                                     adequate to evoke maximal responses 03). E, when the current was terminated during
                                     the response, the response greatly diminished after very little delay. A few miniature
                                     potentials continued for 20--40 msec after the stimulus. F, two examples of expanded
                                     sweeps to show the shape of the responses in detail. The response to the more slowly
                                     rising ramp shows seven individual components clearly. Individual components are not
                                     distinguishable in most of the larger response. Calibrations for A-E are identical.


                             a n d t h e n d e c r e a s e d g r e a t l y , a l t h o u g h u s u a l l y t h e r e c o n t i n u e d t o b e a few
                             m i n i a t u r e p o t e n t i a l s as l o n g as t h e r a m p c u r r e n t w a s m a i n t a i n e d (Fig. 11 C).
                             T h e s a m e d e c r e a s e in responses w a s o b s e r v e d if t h e c u r r e n t w a s h e l d at a
                             c o n s t a n t level slightly a b o v e t h e t h r e s h o l d (Fig. 11 D ) . I f t h e c u r r e n t w a s
                             t e r m i n a t e d , t h e responses s t o p p e d a l m o s t c o m p l e t e l y after a l a t e n c y t o o s h o r t
                             to b e m e a s u r e d a t t h e slow s w e e p s p e e d used (Fig. 11 E).
Published February 1, 1969




                             A. A. AUERBACHAND M. V. L. B~.NNEa"r    ChemicallyTransmitting Synapse   2o5

                                T h e rapid decrease in the response during ramp currents is consistent with
                             the data showing marked depression of transmission following a single spike.
                             If the depression in each case were due to virtually complete depletion of
                             transmitter, it would be predicted that the m a x i m u m amount of transmitter
                             released by a ramp would be more or less equal to that produced by a single
                             spike. In this situation, the response amplitude integrated over time should be
                             approximately the same in the two cases. The data appear to be consistent
                             with this hypothesis, although the comparison is difficult to make accurately
                             because of electrode noise and the "noisy" nature of the responses due to the
                             ramps. T h e time integrals of the largest responses due to the ramps in Fig. 11
                             are about 20 mv msec which is not far from that for a maximal PSP (Fig. 5 E).
                             Furthermore, when a suprathreshold ramp current was followed at varying
                             intervals by a directly evoked spike in the Mauthner fiber, the PSP due to the




                                                                                                                   Downloaded from jgp.rupress.org on May 6, 2011
                             spike was depressed initially and recovered over a time course similar to that
                             observed using a pair of directly evoked spikes. The degree of depression has
                             yet to be correlated with the amount of secretion produced by the ramp,
                             although the depression tended to be greater with larger ramps. T h e simi-
                             larity in the time integrals of response amplitude and in the depressions evoked
                             by ramp currents and spikes suggests that the two forms of depolarization
                             cause secretion by the same mechanism. However, large ramp currents clearly
                             had effects in addition to depletion of transmitter. The ramps had to be given
                             at much lower frequencies than spikes, if the responses were not to become
                             successively more depressed.
                                  DISCUSSION

                             T h e Mauthner fiber, giant fiber synapse is unique among synapses in verte-
                             brate brain in that both pre- and postsynaptic fibers can be penetrated by
                             separate microelectrodes. Several of the transmissional properties are different
                             from those of other known synapses. The PSP is unaffected by current pulses
                             that change presynaptic spike height; depression produced by a single stimu-
                             lus is very pronounced; and maintained depolarization does not cause main-
                             tained transmitter release. All three of these features could be explained by
                             one property, that the immediately available supply of transmitter is easily
                             exhausted. If even a small spike releases all the available transmitter, a large
                             spike can release no more, and PSP's evoked by subsequent spikes will cause
                             little release of transmitter until the immediately available store is replenished.
                             Depletion of transmitter is, of course, not the only explanation of these
                             results. For example, the mechanism of transmitter release could become
                             refractory, or less likely, the postsynaptic membrane could become desen-
                             sitized.
                                Another unusual property is that during high frequencies of stimulation,
                             PSP amplitude is reduced to well below the size of the miniature PSP's seen
Published February 1, 1969




                                    2o6               THE   JOURNAL      OF   GENERAL      ]PHYSIOLOGY      • VOLUME      53   "   I969

                             u n d e r other conditions a n d there are no complete failures of transmission
                             (Fig. 8 F). This finding indicates t h a t either the a m o u n t of transmitter per
                             q u a n t u m is r e d u c e d or the sensitivity of the postsynaptic m e m b r a n e is de-
                             creased; i.e., there is desensitization (24). R e d u c t i o n in q u a n t u m size was
                             observed at the n e u r o m u s c u l a r j u n c t i o n u n d e r conditions where acetylcholine
                             resynthesis was blocked (12), b u t n o t w h e n the nerve was repetitively stimu-
                             lated (10, 13). Repetitive nerve activity does not a p p e a r to cause desensitiza-
                             tion at the n e u r o m u s c u l a r j u n c t i o n (10, 13, 32), b u t p r o b a b l y does do so at a
                             synapse in Aplysia (36). These c o m p a r a t i v e d a t a do n o t lead to a preference for
                             one of the alternative explanations of the effects in the hatchetfish.
                                A critical point in the interpretation of these experiments is the degree to
                             w h i c h the potentials recorded by electrodes in the axonal cores represent the
                             potentials in the terminals. T h e experiments using r a m p currents prove t h a t




                                                                                                                                          Downloaded from jgp.rupress.org on May 6, 2011
                             currents applied in the M a u t h n e r fiber do reach the terminals. T h u s applied
                             currents should affect spike height in the terminals even if the spike a m p l i t u d e
                             differs s o m e w h a t f r o m t h a t recorded in the m a i n t r u n k of the axon. T h e
                             experiments using r a m p currents also validate the experiments showing the
                             existence of a synaptic delay a n d the absence of electrical coupling.

                             Calculations indicate that the decrement in electrotonic spread from axonal core to
                             terminals and from terminals to axonal core is small in both the Mauthner and
                             giant fibers. First, in respect to spread from axonal core to terminals in the Mauthner
                             fiber, the entire input resistance of about 1 M ~ may be ascribed to eight contra-
                             and two ipsilateral terminals, each of which then has an input resistance of 10 Mf~.
                             This value is likely to be much too low because the space constant is so long (Fig. 4)
                             that the resistance measurements must involve spread to nodes distant from the
                             synaptic region. If each terminal has a 20 ~ long, 5/z diameter myelinated portion
                             whose surface resistivity is infinite and whose axoplasmic resistivity is 100 f~ cm, the
                             intracellular access resistance to the terminal's unmyelinated portion is about 1 Mfl.
                             As the total input resistance of each terminal exceeds 10 Mf], only a small voltage
                             drop could occur in the myelinated part of the terminal. If the unmyelinated portion
                             of each terminal has a total surface area of 200 #~, and if an input resistance of 9 Mf~
                             is ascribed to this membrane, the calculated membrane resistivity is 18 fl cm ~. Again,
                             this value is likely to be a marked underestimation because of the long space con-
                             stants. If one assumes a fiber 5/z in diameter with 18 f] cm 2 membrane resistivity and
                             axoplasmic resistivity of 100 f~ cm, the space constant would be greater than 45/~.
                             Since the unmyelinated portion of the terminals is considerably shorter than 45/~
                             (Fig. 1 C and D) and since it constitutes a core conductor with a closed end, there
                             can be little decrement within the unmyelinated part of the terminals. It can be
                             concluded that a large part of the potential recorded in the axonal core due to ap-
                             plied current is developed across the membranes of the terminals. The degree of
                             nonisopotentiality could be greater during a spike because of membrane capacity
                             and increased conductance during activity. However, spikes would not be expected
                             to decrement significantly in inactive terminals. The time constant of the terminal
Published February 1, 1969




                             A. A. Atmm~ACH AND M. V. L. BENN]STT ChemicallyTransmittingSynapse                               2o7

                             membrane would be very short compared to spike duration if its resistance were as
                             low as 10 fl em 2 and its capacity were I # F / c m ~. If its resistivity were higher, there
                             would be even less decrement in the terminal although the membrane time constant
                             would be greater. Assuming that the terminals generate spikes, a lO-fold decrease in
                             resistance would lead to about a 3-fold decrease in space constant, and the spike at
                             the tip of a terminal might differ somewhat from that in the axonal core. However,
                             it would be difficult to explain the postactivation depression by failure of impulses
                             to propagate into the terminals, because there is no change in the Mauthner fiber
                             spike over most of the period of depression. Furthermore, the PSP's are little affected
                             by marked changes in spike height.
                                In respect to current spread from terminal to axonal core in the Mauthner fiber,
                             there are two limiting cases. If all the input resistance is ascribed to the terminals,
                             there can be no decrement in the myelinated portion of the terminals, and, as before,
                             negligible decrement in the unmyelinated portion. If on the other hand, the terminal




                                                                                                                                             Downloaded from jgp.rupress.org on May 6, 2011
                             membrane is assumed to be of infinite resistance, the total access resistance to the
                             core provided by four myelinated terminals would be 0.2 Mf~, a small fraction of the
                             fiber's input resistance. One can eonclude that any electrotonic spread from giant
                             fiber to ipsilateral Mauthner fiber would be little affected by loss in the Mauthner
                             fiber terminals. Electrotonic spread from giant fiber to contralateral Mauthner fiber
                             would show some decrement along the single Mauthner fiber terminal connecting
                             them, but should nevertheless be detectable if the PSP's were electrically transmitted.
                                In respect to the giant fibers, the input resistance is about 0.5 Mf~ but the resistance
                             ascribable to processes synapsing with the Mauthner fiber is greater than 1 M r
                             because of spread into the motoneurons (3). Furthermore, the giant fiber processes
                             are considerably shorter and thicker than the Mauthner fibers terminals (Fig. 1
                             C, D). Thus, there is likely to be even less decrement than is indicated for the Mauth-
                             ner fiber. In any case, attenuation of PSP's in spreading from terminal into axon
                             would not affect the observation of reduction of PSP amplitude during high frequency
                             stimulation to below that of the normal miniature PSP's.

                             I f one accepts the hypothesis t h a t the i m m e d i a t e l y available t r a n s m i t t e r is
                             largely e x h a u s t e d b y a single spike, the p r o b a b i l i t y of release b y a spike is high
                             for e a c h q u a n t u m in this store. A t a n u m b e r of synapses, the a m p l i t u d e s of the
                             PSP's are described b y a Poisson distribution (29). T h i s t y p e of distribution
                             is usually ascribed to a process in w h i c h a small p r o b a b i l i t y of release of e a c h
                             q u a n t u m operates on a large n u m b e r of q u a n t a . A t the hatchetfish synapse,
                             the steady-state a m p l i t u d e variations at m o d e r a t e frequencies of stimulation
                             d o n o t a p p e a r to fit a Poisson distribution because the coefficient of v a r i a t i o n
                             is too small. A t t e n u a t i o n of PSP's in spreading f r o m terminals to a x o n a l core
                             c o u l d n o t explain the d e v i a t i o n f r o m a Poisson distribution. T h e r e w o u l d b e no
                             effect on the m e a s u r e d distribution if the a t t e n u a t i o n w e r e the s a m e a t e a c h
                             t e r m i n a l ; if the a t t e n u a t i o n differed at different terminals, t h e r e w o u l d be a n
                             i n c r e a s e in the coefficient of variation. A l t h o u g h f u r t h e r s t u d y is r e q u i r e d ,
                             the v a r i a n c e of the P S P a m p l i t u d e distribution differs f r o m t h a t of a Poisson
Published February 1, 1969




                                  2o8           THE   JOURNAL   OF   GENERAL   PHYSIOLOGY   • VOLUME   53   "   I969

                             distribution in the direction expected for a binomial distribution of amplitudes
                             where the probability of quantal release is high. The magnitude of the coeffi-
                             cient of variation appears consistent with the prediction from a binomial
                             distribution in which the probability of quantal release is that which was
                             measured by the degree of depression using paired stimulation (Fig. 7 A).
                                 It should be noted that a large probability of quantal release does not
                             require that the amplitudes of the PSP's have a non-Poisson distribution, since
                             the output can also reflect the statistics describing the immediately available
                             store. Thus, if a process having a large (or small) probability of release
                             operates on an immediately available store whose size varies according to a
                             Poisson distribution, the output of transmitter will be distributed according to
                             Poisson statistics (35). This fact is obvious for the limiting case of release
                             probability equal to one. To give a physical picture, let the immediately




                                                                                                                       Downloaded from jgp.rupress.org on May 6, 2011
                             available store be the number of vesicles occupying release sites at a given
                             instant. This number would have a Poisson distribution if vesicles containing
                             quanta were freely diffusing near many release sites, each of which had only
                              a low probability of being occupied by a vesicle. In these circumstances,
                             output would be Poisson-distributed whether probability of quantal release
                             were low or high. If the probability that a release site were occupied and the
                             probability of quantal release were both large, the amplitude distribution
                             would become binomial (35). Restricted diffusion of vesicles within the ter-
                             minal and replenishment of the immediately available store might also lead
                             to a nonPoisson distribution of PSP amplitudes. In the hatchetfish the latter
                             factor could be important in the brief period of greater variability observed
                             shortly after the onset of a stimulus train (Fig. 7 B).
                                If release probability were in fact high, the size of a PSP would be a measure
                             of the size of the immediately available store of transmitter, which could be
                             little more than the total mobilized since the previous stimulus. A figure for
                             the number of quanta in the immediately available store can be estimated
                             from quantal size, about 0.4 mv, and PSP amplitude, 20-40 mv. The value
                             of 50-100 quanta is somewhat smaller than those given for neuromuscular
                             junctions, but comparable to those for the synpathetic and ciliary ganglia (9,
                             29, 31). The decrease in PSP size produced by a short period of high frequency
                             stimulation suggests that the total amount of mobilizable transmitter may also
                             be small. Study of the time course of amplitude changes during and after
                             tetani should make it possible to estimate the size of the mobilizable store as
                             well as the rate at which it is refilled, perhaps by resynthesis, and the rate at
                             which it is emptied by movement of transmitter into the immediately available
                             store. Analysis of variance of PSP amplitudes may also be useful in defining
                             the changes that occur as a result of tetanic stimulation.
                                Further study of this synapse should prove valuable in elucidating the
                             mechanism of chemically mediated transmission and the relation between
Published February 1, 1969




                             A. A. AUEI~ACH AND M. V. L. BENNETT Chemically Transmitting Synapse                                                  209

                             p r e s y n a p t i c p o t e n t i a l a n d r e l e a s e o f t r a n s m i t t e r . A l t h o u g h t h i s s y n a p s e is
                             d i f f e r e n t in s e v e r a l r e s p e c t s f r o m o t h e r s s t u d i e d t o d a t e , i t m a y w e l l b e r e p r e -
                             sentative of many synapses in the central nervous system.

                             The earlier work was presented to the Department of Biophysics, Columbia University, by A. A.
                             Auerbach in partial fulfillment of the requirements for the Ph.D. degree.
                             We are indebted to Mr. Sidney Steinberg and Mr. Victor Klieg of the Interdisciplinary Electronics
                             Facility of the Albert Einstein College of Medicine for their assistance in the design and construc-
                             tion of the ramp current stimulator and other electronic equipment.
                             This investigation was supported in part by grants from the National Institutes of Health (a Post-
                             doctoral Fellowship under NIH 5TI-MH 6418-11; Career Program Award K3-GM-5828; Public
                             Health Service Research Grants NB-3728, NB-3270, 5TI-NB-5328, NB-03448, NB-03313, and
                             NIH 1 P01-NB07512-01), from the National Science Foundation (GB 2940 and GB 6880), and
                             from the U.S. Air Force (AFOSR 550).
                             Receivedfor publication 20 June 1968.




                                                                                                                                                                    Downloaded from jgp.rupress.org on May 6, 2011
                                          REFERENCES
                               1. AUERBACH,A. A., and M. V. L. BENNETT. 1967. Chemically and electrically transmitting
                                    junctions in the central nervous system of the hatchetfish, Gasteropdeeus. J. Gen. Physiol.
                                    50:1090.
                               2. Atmm3ACH, A. A., and M. V. L. BENNV.TT. 1968. Chemically mediated transmission at a
                                     giant fiber synapse in vertebrate brain. Proceedings 24th International Congress Physi-
                                     ological Sciences, Washington, D. C. 7: 21.
                               3. AtrE~ACH, A. A., and M. V. L. BENNEa'r. 1969. A rectifying electrical synapse in the
                                     central nervous system of a vertebrate. J. Gen. Physiol. 53:21 I.
                               4. BENNETT, M. V. L. 1964. Nervous function at the cellular level. Annu. Rev. Physiol. 26: 289.
                               5. BENNETT, M. V. L. 1966. Physiology of electrotonic junctions. Ann. N.Y. Amd. Sd. 137:
                                     509.
                               6. BENNETT, M. V. L. 1968. Similarities between electrically and chemically mediated trans-
                                     mission. In Physiological and Biophysical Aspects of Nervous Integration. F. D. Carl-
                                     son, editor. Prentice-Hall, Inc., Englewood Cliffs, N . J . p . 73.
                               7. BENN~.TT,M. V. L., M. G~.NgZ, Y. NAKAJ~A, and G. D. PAPPAS. 1967. Physiology and
                                     ultrastructure of electrotonic junctions. IV. Medullary electromotor nuclei in gymnotid
                                     fish. J. Neurophysiol. 30: 236.
                               8. BENN~Tr, M. V. L., Y. NAXAJ~A, and G. D. P~a,X~AS.1967. Physiology and ultrastructure
                                     of electrotonic junctions. III. The giant electromotor neurons of Malapterurus eleari-
                                     cus. J. Neurophysiol. 30: 209.
                               9. BLACr~UAN,J. G., B. L. Gn~SBORQ, and C. RAY. 1963. On the quantal release of the trans-
                                     mitter at a sympathetic synapse. J. Physiol. (London). 167: 402.
                              I0. BRoom, V. B., and R. E. Trtms. 1962. Reduction of quantum content during neuromus-
                                     cular transmission. J. Physiol. (London). 162: 298.
                              11. EIOENMANN,C. H. 1912. The fresh water fishes of British Guiana. Mem. Carnegie Museum.
                                     5: 1-554.
                              12. ELMQtnST, D., and D. M. J. QUASa~L. 1965. Presynaptic action of hemicholinium at the
                                     neuromuscular junction. J. Physiol. (London). 177: 463.
                              13. ELMQtnST, D., and D. M. J. Q.UASa~L. 1965. A quantitative study of end plate potentials
                                     in isolated human muscle. 3". Physiol. (London). 178: 505.
                              14. FRANK, K., and M. G. F. FuoRa~s. 1956. Stimulation of spinal motoneurons with intra-
                                     cellular electrodes, or. Physiol. (London). 134:45 I.
                              15. FUgSt-WAN,E. J., and D. D. POTa~R. 1959. Transmission at the giant motor synapses of
                                     the crayfish, or. Physiol. (London). 145: 289.
                              16. GRUnI~FEST, H. 1961. General physiology and pharmacology of junctional transmission.
Published February 1, 1969




                                     210              THE    JOURNAL        OF   GENERAL   PHYSIOLOGY   • VOLUME    53   "   1969
                                    In Biophysics of Physiological and Pharmacological Actions. A. M. Shanes, editor.
                                    American Association for the Advancement of Science, Washington, D . C . p . 329.
                             17.   HAGIWARA, S., and I. TASAKL 1958. A study of the mechanism of impulse transmission
                                     across the giant synapse of the squid. J. Physiol. (London). 143" 114.
                             18.   KATZ, B. 1962. The transmission of impulses from nerve to muscle, and the subcellular
                                    unit of synaptic action. Proc. Roy Soc. Ser. B. Biol. Sci. 155: 455.
                             19.   KATZ, B., and R. MILEDI. 1965. The effect of temperature on the synaptic delay at the
                                    neuromuscular junction. J. Physiol. (London). 181: 656.
                             90.   KATZ, B., and R. MILEDI. 1965. The measurement of synaptic delay and the time course
                                    of acetylcholine release from motor nerve terminals. Proc. Roy. Soc. Set. B. Biol. Sd. 161:
                                    496.
                             21.   KATZ, B., and R. MmEDI. 1967. Tetrodotoxin and neuromuscular transmission. Proc.
                                     Roy. Soc. Ser. B. Biol. Sci. 167: 8.
                             22. KATZ, B., and R. MIL~.DL 1967. The release of acetylcholine from nerve endings by graded
                                   electric pulses. Proc. Roy. Soc. Ser. B. Biol. Sd. 167: 23.
                             23. KATZ, B., and R. Mn.EDL 1967. A study of synaptic transmission in the absence of im-
                                   pulses. J. Physiol. (London). 192: 407.




                                                                                                                                    Downloaded from jgp.rupress.org on May 6, 2011
                             24. KATZ, B., and S. Tm~SL~.PF. 1957. A study of "desensitization" produced by acetylcholine
                                   at the motor end plate, dr. Physiol. (London). 138: 63.
                             25. KEYNES, R. D., and H. MART~S-FERm~mA. 1953. Membrane potentials in the electro-
                                   plates of the electric eel. J. Physiol. (London). 119: 315.
                             26. KuNo, M. 1964. Quantal components of excitatory synaptic potentials in spinal moto-
                                   neurons. J. Physiol. (London). 175: 81.
                             27. KUSANO,K. 1968. Further study of the relationship between pre- and postsynaptic poten-
                                   tials in the squid giant synapse. ,7. Gen. Physiol. 52: 326.
                             28. KUSANO,K., D. R. LIWNOOOD, and R. WERMAN. 1967. Correlation of transmitter release
                                   with membrane properties of the presynaptic fiber of the squid giant synapse, or. Gen.
                                   Physiol. 50: 2579.
                             29. MARTIN, A. R. 1966. Quantal nature of synaptic transmission. Physiol. Rev. 46: 51.
                             30. MARTIN, A. R., and G. PILAR. 1964. Presynaptic and postsynaptic events during post-
                                   tetanic potentiation and facilitation in the avian ciliary ganglion. 3". Physiol. (London).
                                   175: 17.
                             31. MARTIN, A. R., and G. PmAR. 1964. Quantal components of the synaptic potential in the
                                   ciliary ganglion of the chick. J. Physiol. (London). 175: 1.
                             32. O~tmA, M., M. ENDO, and Y. NONOMURA. 1962. Presynaptic nature of neuromuscular
                                   depression. ,lap. J. Physiol. 12-" 573.
                             33. TaxEucm, A., and N. T~,.~ucm. 1969. Electrical changes in pre- and postsynaptic axons
                                   of the giant synapse of Loligo. ,7. Gen. PhysioL 45:1181.
                             34. THOMAS, R. C., and V. J. WIt.SON. 1966. Marking single neurons by staining with intra-
                                   cellular recording microelectrodes. Science. 151:1538.
                             35. VE~-JoNES, I3. 1966. Simple stochastic models for the release of quanta of transmitter
                                   from a nerve terminal. Aust. J. Statist. 8: 53.
                             36. WACHa~., H., and E. R. KANDEL. 1967. A direct synaptic connection mediating both
                                   excitation and inhibition. Science. 158: 1206.
                             37. WATANABE,A., and H. GRtrNDFEST. 1961. Impulse propagation at the septal and com-
                                   missural junctions of crayfish. J. Gen. Physiol. 45: 267.
                             38. WEITZ~AN, S. H. 1954. The osteology and relationships of the South American characid
                                   fishes of the subfamily. Gasteropelindnae. Stanford lchthyol. Bull. 4.

				
hkksew3563rd hkksew3563rd http://
About