the organization of synaptic axoplasm in the lamprey … · the filial filaments described by...

THE ORGANIZATION OF SYNAPTIC

AXOPLASM IN THE LAMPREY (PETROMYZON

MARINUS) CENTRAL NERVOUS SYSTEM

INTRODUCTION

The cyclostomes, lampreys and hagfish, representthe most primitive living vertebrates, a positionthat singles them out as of special interest in studiesof the evolution and comparative physiology of thenervous system. The preliminary electron micro-scopic studies of Schultz, Berkowitz and Pease(1956) and the later work of Bertolini (1964) es-tablished several fine structural features of thecyclostome nerve cord that reflect the phyloge-netic position of these animals. Bertolini points outthat these features include the dorsoventrallyflattened form of the cord, the horizontal disposi-tion of the gray matter, and the presence in thecord of giant fibers arising from cell bodies withinthe brain. Furthermore, the lamprey nerve cordshares with that of an insect (Wigglesworth, 1960)

DAVID S . SMITH, U . JARLFORS, and R . BERANEK

From the Papanicolaou Cancer Research Institute and the Department of Medicine, School ofMedicine, University of Miami, Miami, Florida 33152

ABSTRACT

The fine structure of synapses in the central nervous system of lamprey (Petromyzon marinas)ammocoetes has been investigated . Both synapses within the neuropil and synaptic linksbetween giant fibers (including Muller cells) and small postsynaptic units are described .The distribution of neurofilaments and microtubules in nerve profiles over a wide diameterrange is described, and the possible role of these structures in intracellular transport isdiscussed . Electron micrographs indicate that small lucent "synaptic vesicles" occur sparselythroughout the axoplasm and in regular arrays in association with microtubules in thevicinity of synapses. Within a synaptic focus, immediately adjoining the presynaptic mem-brane, vesicles are randomly arranged and are not associated with microtubules . Neuro-filaments are present, generally in large numbers, but these are not associated with vesiclesor other particulates . The structural findings are considered in terms of current conceptsof fast and slow transport in neurons and the mechanochemical control of intracellularmovement of materials .

the absence of an intramedullary blood supply . Ithas also been established that the `white matter' ofthe cyclostome cord contains no myelin, andBertolini has described some aspects of neuronaland glial relations in the central nervous system,including some morphological features of centralsynapses .

A large number of electron microscopic studieson the central and peripheral nervous system ofvertebrates and invertebrates have shown thatwhile the neuron cell body contains the nucleusand a varied complement of intracellular compo-nents, the axon processes, and dendrites whenpresent, are organized in a characteristic thoughstructurally less complex fashion. The axoplasm ofperipheral axons and nonsynaptic regions of the

THE JOURNAL OF CELL BIOLOGY . VOLUME 46, 1970 . pages 199-219

199

Dow

nloaded from http://rupress.org/jcb/article-pdf/46/2/199/1263354/199.pdf by guest on 08 Septem

ber 2021

central nerve cord contain longitudinally orientedmicrotubules similar in appearance to those de-scribed in very diverse situations in animal andplant cells . The microtubules are accompanied, invarying ratios, by similarly oriented but smaller"neurofilaments ." Mitochondria accompany theseaxoplasmic structures, but generally not in im-pressive numbers. Points of chemically mediatedsynapse, whether in the central or peripheral re-gions and irrespective of the chemical nature ofthe transmitter, are associated with focal concen-trations of small vesicles (De Robertis, 1958 ; Palay,1958 ; Katz, 1962 ; Whittaker, 1968) . An attractivehypothesis of synaptic action involves the synthesisand/or storage of transmitter molecules in thesevesicles and the subsequent release of transmitterinto the synaptic gap upon arrival of excitation atthe presynaptic terminal . However, the mechanismby which transmitter molecules are transported toor synthesized at the ending is incompletely under-stood but represents a crucial aspect of nervefunction .

Recent studies have revealed the operation ofdistinct rapid and slow translocation of materialswithin the axon, perhaps including materials in-volved in synaptic function . There is at presentmuch interest in the physical pathways availablefor intracellular transport of particulate and othercomponents, and the form of the neuron togetherwith the fact that the axoplasm is relatively simple,but includes highly ordered structures, singles thiscell out as unusually convenient for such studies .The present work forms part of a general investi-gation of the central nervous system of the Petro-myzon ammocoete and provides morphological evi-dence for an association between vesicles and alinear axoplasmic component that is preciselyzoned with respect to the presynaptic surface . Thepossible functional implications of these findingsare discussed .

MATERIALS AND METHODS

Ammocoete larvae of Petromyzon marinus, about12-14 cm in length, were used in this study. Thelampreys were obtained from the Ocqueoc River ofLake Huron and kindly provided by Mr. LouisKing, Acting Investigation Chief, Bureau of Com-mercial Fisheries, Hammond Bay Biological Station,Millersburgh, Michigan . Animals were shipped byAir Express to Miami and kept in shallow tanks ofaerated water in a 4 °C room. Animals were de-capitated and a length of the body caudal to theorigin of the first dorsal fin was separated with a

200

THE JOURNAL OF CELL BIOLOGY . VOLUME 46, 1970

razor blade . Chilled fixative (2 .5% glutaraldehydein 0 .05 M cacodylate buffer at pH 7 .4, containing1 % sucrose) was gently infused into the perineuralspace surrounding the spinal cord from a hypo-dermic syringe. Immediately after, the cord wasexposed by cutting into the body wall and strippingoff the dorsal meninges; the cord was removed asgently as possible and placed in a vial of chilledfixative for 4 hr. The material was next washedovernight in chilled cacodylate buffer containing2% sucrose, and at this stage the cord was cut into2 mm lengths. The segments were placed in chilled1 % Os04 in the same buffer for 1 hr, dehydrated inan ethanol series commencing at 50%, and em-bedded in Araldite (Ciba Ltd ., Duxford, Cam-bridge, England) . Sections were cut with glassknives on an LKB Ultrotome III (LKB Instruments,Inc ., Rockville, Md.) collected on uncoated grids,stained in 50% ethanolic uranyl acetate (saturated)and lead citrate, and examined in a Philips EM 200 .For light microscopy, 1 p. sections of the Araldite-embedded material were stained in hot 1 % toluidineblue with 1 % borax, and examined in a Zeiss micro-scope equipped with planapochromatic objectives .

RESULTS

The general organization of the Petromyzon ammo-coete spinal cord is illustrated in a light micro-graph in Fig . 1 . It is dorsoventrally compressed,and on either side of the neural canal the graymatter, containing neuron cell bodies, extends asa horizontally disposed flange . The rest of the cordcontains glial cells and processes together withnerve axons and dendrites, ranging from extremelysmall components beyond the resolution of thelight microscope to large fibers (of the Mauthnerand Müller cells) which, in the larvae used here,reach a diameter of circa 30 µ . Fig. 2 includes agroup ofMüller fibers at higher magnification : thesignificance of details resolved in such a prepara-tion only become fully apparent when correspond-ing electron micrographs are examined . The smalldots scattered throughout the axoplasm representfiliform mitochondria aligned with the long axis ofthe cell, while the densely stained patches adjoin-ing the axon surface correspond to the clusteredvesicles of the synaptic foci (cf . Fig . 5) . The resolu-tion of tight aggregates of synaptic vesicles in thelight microscope does not appear to have beenreported previously, and may perhaps permit moreextensive histochemical analysis of synapses insuch specialized neurons as these .

Except for the complete absence of myelin, re-gions of the cord which may conventionally be

Dow


ber 2021

described as the "white matter" display fine struc-tural features shared with comparable regions ofother chordate nervous systems, together withsome primitive aspects. Fig . 3 illustrates an elec-tron micrograph of a transversely sectioned groupof axons adjoining the dorsal surface of the cord .These are divided into tracts by glial sheets which,as elsewhere in the cord, are replete with cyto-plasmic filaments. The general features of axon-glia relations have been described by Schultz et al .(1956) and Bertolini (1964) who confirmedRetzius' (1891) supposition that the cyclostomespinal cord includes only a single type of glial cell .The arrangement of structures within the axoplasmof nonsynaptic areas, on the other hand, presentsno obviously unusual features. As illustrated inFig. 4, both tubular and filamentous structuresoccur in the axoplasm . Schmitt and Samson (1969)have reviewed the distribution of microtubules andneurofilaments in nerve processes . With the excep-tions described by Wuerker and Palay (1969),these components occur in an approximately re-ciprocal balance-in general, small axons possess-ing numerous tubules but few filaments, with thereverse holding true in large axons. This generalpattern is encountered in the nerve cord ofPetromyzon where in the material used here thediameter of nerve processes ranges from circa 30 to0 .1 it and (Bertolini, 1964) up to 100 is in adults.In the smallest processes observed here, three orfour microtubules may occur together in the ab-sence of filaments ; about 10 of each are presentwhen the diameter of the process is of the order of0.3 Ä . Larger profiles in the range of 1-2 it haveneurofilament :microtubule ratios in the range of4 :1-10:1, while the ratio in axoplasm of the giantfibers is circa 20-30 : 1 . As illustrated in Fig . 4, themicrotubules occur singly or more rarely infascicles of up to six or seven in which a minimumintertubule distance of 100 A is maintained andwhich appear to include linkages between thetubules. As described in some other neurons(Gonatas and Robbins, 1965 ; Echandia et al .,1968), transverse profiles frequently possess a densecore or "dot," circa 50 A in diameter. The latterauthors have shown in longitudinal profiles thatdense material is discontinuously distributed alongthe axis of the tubule, apparently accounting forthe sporadic occurrence of the core in transversesections. However, Lane and Treherne (1969), onthe other hand, have stained the entire micro-tubule core with lanthanum nitrate, suggesting

that it is uniformly accessible to small molecules .A blank space or halo surrounds each tubule orgroup, as is usually the case, and the remainderof the axoplasm contains uniformly distributedneurofilaments and occasional mitochondria .Filaments circa 100 A in diameter have been de-scribed as a special feature of neurons, and withinthese has been resolved a 30 A electron-lucent core(Palay, 1964 ; Wuerker and Palay, 1969 ; Schmittand Samson, 1969 ; Wuerker, 1970). In lampreyneurons, the neurofilaments appear to be signifi-cantly smaller than elsewhere (Fig . 4) ; their over-all diameter is only 60 A . These resemble in sizethe filial filaments described by Wuerker (1970),but unlike these and in common with otherneurofilaments, wispy cross-bridges are present inthe lamprey axoplasm extending between fila-ments, which have a minimum intervening spaceof 300 A. In Petromyzon, the conspicuous glial fila-ments are circa 60 A in diameter (Fig . 11), and theminimum interfilament space is likewise circa 60 A .Fig. 5 illustrates a transversely sectioned field in-cluding a small portion of the peripheral axoplasmof a giant fiber (cf. Figs . 1, 2), probably the axonof a MUller interneuron (Rovainen, 1967) linkinginput from a variety of sensory sources to motorunits controlling muscle movements . These fibersengage in profuse and extremely localized syn-apses with small postsynaptic units, probablydendrites of motor neurons. As Bertolini noted, thesynapses are demarcated by the localized groupingof synaptic vesicles at the periphery of the axon .Two synaptic foci are included in Fig. 5, illus-trating with almost diagrammatic clarity thestructural features recognized (Palay, 1958 ;De Robertis, 1958, 1967; and others) as associatedwith sites of chemically mediated synapse . Clustersof small lucent vesicles are grouped around a circa0.5 is length where the plasma membranes of pre-synaptic (giant) cell and small postsynaptic mem-ber are juxtaposed across a synaptic gap circa 150 Ain width . This field also includes two groups ofsynaptic vesicles stopping just short of the plasmamembrane and approaching their points ofsynapse which lie a few tenths of a micron out ofthe plane of this section . Examination of serial sec-tions (600 A in thickness) suggests that areas ofclose membrane apposition are approximatelycircular with a diameter of circa 0.5,u .

In myoneural junctions, synaptic vesicles aregrouped into axon endings that are small by com-parison with the innervated cell, and the same is

DAVID S . SMITH, U. JARLFORS, AND R . BERnNEK Synaptic Axoplasm in Petromyzon

201

Dow


ber 2021

2 0 2

Dow


ber 2021

true of synapses onto nerve cell bodies . In Petro-myzon giant fibers, on the other hand, a very dif-ferent situation exists . The presynaptic axon sur-face is a mosaic : a cylinder that is associated witha very large number of discrete punctate synapses .In proportion to the entire cell, these vesicleclusters merely represent small irregularities in thesmooth axon cylinder, while the bulk of the axo-plasm contains great numbers of neurofilaments(circa 250-350/Äe) and 20 to 30 times fewer micro-tubules . The disposition of these oriented struc-tures with respect to the synapse is seen at highermagnification : Fig. 6 includes a pair of almost con-fluent MUller cell synapses, involving sphericalsynaptic vesicles 450-550 A in diameter, in a con-centration of about 3000-40O0/Ä 3 . This field alsoincludes the synaptic cleft and the often noted"thickening" particularly of the postsynaptic mem-brane, which in this instance results from two quiteseparate structures : (a) wispy dense material liningthe cytoplasmic surface of the postsynaptic cell,and (b) particles projecting into the synaptic cleftfrom the postsynaptic cell membrane .

Similar groups of synaptic vesicles occur pro-fusely throughout the neuropil regions of thePetromyzon spinal cord, linking innumerable smallerand as yet unidentified nerve processes . In addi-tion, as elsewhere in vertebrate central nervoussystems and in sharp distinction to arthropodcentral ganglia, many processes terminate on cellbodies within the gray matter .

In view of the possibility that axoplasmic struc-tures may be involved in transport of materialsfrom the perikaryon, perhaps including materialsdestined from synaptic areas, it is of interest toestablish whether neurofilaments and/or neuronal

microtubules show any special features such aspreferential location in the vicinity of synapses .Counts of these structures in giant fibers andsmaller profiles lead to the following conclusions .(a) Neurofilaments are uniformly distributed vir-tually throughout the neuron profile except in thesmallest profiles where they are not numerous and,except within the clusters of synaptic vesicles,where they appear to be absent . (b) Microtubulesare not strikingly concentrated within synapticfoci, occurring at a concentration of circa 10-20/Äzboth in these regions and elsewhere in profiles ofgiant and medium-sized axons in Petromyzon . How-ever, despite this apparent lack of "recognition" ofsynaptic foci by linear axoplasmic structures, thesynaptic vesicles display an extremely precise andspecific affinity for microtubules in a restrictedzone of the synaptic focus, described in a prelimi-nary account (Jarlfors and Smith, 1969) and hereillustrated further in Fig . 7. This micrograph in-cludes numerous synaptic vesicles within an axoncirca I Ä in diameter, but illustrates a configurationthat occurs equally in the largest fibers of theammocoete nerve cord (cf. Fig. 6) . Randomlyscattered vesicles are accompanied by petal-likeclusters of vesicles, but incomplete groups fre-quently occur. The crispness of the vesicle profileswithin a rosette is often variable, suggesting thatthe vesicles are not simply arranged in a planetransverse to the long axis. Moreover, the rosettepattern is slightly asymmetric ; adjacent vesiclesare either separated by a narrow gap or show slightoverlap. The center-to-center distance between thevesicles and the microtubule core is circa 400 A,and vesicles are separated from the tubule surfaceby a space of 80-120 A, in which no structure has

FIGURE 1 Light micrograph of a transversely sectioned Petromyzon ammocoete nerve cord embeddedin Araldite after processing for electron microscopy, cut at a thickness of 1 Ä, and stained with toluidineblue. The medial neural canal is indicated by an arrow . A narrow flange of gray matter (white asterisks)occupies the central region of the ribbon-like cordon either side of the neural canal (arrow) ; this containsneurone cell bodies, and the bulk of the cord contains small nerve processes (black asterisks) correspond-ing to the white matter of higher chordate nerve cords, though as noted by Schultz et al . (1965) theselack myelin . Large axons (ax) include the giant Muller fibers . X 750 .

FIGURE 2 A section similar to Fig . 1 at higher magnification . Cell bodies of small neurons in the graymatter are included (*) beneath which lie axon profiles ranging from the limits of the light microscopeto large cells (ax) circa 25 Ä in diameter, representing the `giant' Muller fibers . The details of axoplasmicorganization in such preparations become clear only after examination of electron micrographs, as inFig . 5 . The small dots in the axoplasm (arrowheads) evidently represent the filiform mitochondria . Moreinterestingly, the discrete heavily stained patches, each about 1 Ä in width (arrows) adjoining the axonsurfaces, correspond to the compact foci of synaptic vesicles observed in electron micrographs (cf . Figs .5 and 6) . X 2000.

DAVID S . SMITH, U. .IXRLFORS, AND R. BERANEK Synaptic Axoptasm in Petromyzon

203

Dow


ber 2021

204

Dow


ber 2021

yet been resolved . No evidence has been found oflateral bridges or other linkage devices (such asthe neurofilament side arms) between the vesiclesand tubules.

In longitudinal aspect, this association is stillmore striking . The low-power survey field shownin Fig . 8 illustrates an unusually extensive align-ment of synaptic vesicles on a framework of micro-tubules, over a distance of 5 Ä. A similar configu-ration is shown at higher magnification in Fig. 9,in which a tight column of electron-lucent vesiclesis arranged on a linear tubular scaffolding adjoin-ing, at a short distance, a virtually unaccompaniedtubule. Transverse sections (Figs . 6 and 7) indi-cated that the vesicle-to-tubule association is agenerally regular one, and longitudinal profilesreveal that the petal-like grouping results from theclose packing of vesicles around microtubule cores .

The vesicle-clad tubules in Figs . 8 and 9 maywell adjoin synaptic areas, but if so the apposedpre- and postsynaptic membranes lie out of theplane of section . The restriction of microtubules tothe general axoplasm and the outer zone of thesynaptic focus and their absence among the vesiclesimmediately adjoining the presynaptic membrane(Fig. 6) are general features of transverse profiles,and this disposition is also evident in suitablelongitudinal sections . In Fig . 10 a vesicle-cladtubule passes within % ,u of a point of synapse : inthis % to zone microtubules are absent, and ran-domly arranged and unassociated vesicles clusteraround the presynaptic membrane (cf. Fig. 6) .Similarly, in Fig. 11 a large axon adjoins a smallpostsynaptic unit; vesicles are tightly groupedaround microtubules (as in Fig. 9) some distancefrom the apposed synaptic membranes, but closerto the latter tubules are associated with vesiclesonly to the level of the synapse . In Fig. 11, un-attached vesicles extend to the localized patch of

the axon surface thought to be site of transmitterrelease, while below this level in the micrographextend naked microtubules .

The synaptic vesicles that have been illustratedin the accompanying electron micrographs areuniformly spherical and electron lucent . Beforethis section is concluded, two further axoplasmiccomponents that have been noted in the ammo-coete spinal cord must be mentioned . (a) A largervesicle varying widely in diameter and containinga denser pellet or core is present consistently, butinfrequently, in populations of the smaller lucentvesicles . In Figs . 12 and 13, the size range of thiscomponent is 700-1200 A, though slightly smallerand larger examples have been noted . These vesi-cles do not seem to be preferentially located nearthe presynaptic surface, nor have they been notedin the columns of vesicles surrounding the micro-tubules . (b) Each of the major structural featuresof ammocoete synapses described so far is dupli-cated in central axon profiles smaller than thegiant fibers, and distinguishable by the shape ofits synaptic vesicles . Several authors (Atwood andJones, 1967 ; Uchizono, 1965, 1967 ; Nakajima,1970; and others) have reported ellipsoidal or dis-coidal synaptic vesicles in preparations that alsoinclude neurons containing the spherical type . Ithas been suggested by these authors that by afortunate chance, from the point of view of inter-pretation of electron micrographs, the flattenedvesicles may be associated with inhibitory ratherthan excitatory endings . Similar flattened vesiclesoccur frequently in thin sections of the Petromyzonspinal cord. Their distribution relative to thespherical vesicles in this material will be describedin detail elsewhere (Smith and Beranek, manu-script in preparation) but it may be mentioned insummary that an axon appears never to contain amixture of lucent spherical and nonspherical vesi-

FIGURE 3 A group of axons (ax) within the central nerve of a Petromyzon marinas ammocoete . Thetransverse nerve profiles are separated into tracts by sheets of fibrous glial cytoplasm (g) . As in othernonsynaptic regions the axoplasm contains microtubules, neurofilaments (seen at higher magnificationin the next figure), and occasional mitochondria . X 21,000.

FIGURE 4 Illustrating typical nonsynaptic axoplasm of a Petromyzon central axon in more detail thanin Fig. 1 . The principal structural components are microtubules, circa 250 A in diameter, and neurofila-ments, circa 60 A in diameter . The former occur single (t) and in compact groups (t') with a center-to-center spacing of circa 400 A. Microtubules frequently exhibit an opaque core (lower insert) . Neurofila-ments (f) are situated in the surrounding axoplasm ; they appear to be linked by side arms (arrows), andas illustrated in the upper insert, the filament includes an electron-lucent core . X 100,000 ; inserts,X 370,000 .

DAVID S. SMITH, U . JARLFORS, AND R . BERANEK Synaptic Axoplasm in Petromyzon

205

Dow


ber 2021

2 0 6

Dow


ber 2021

cles, and that the arrangement of vesicles withrespect to axoplasmic microtubules and synapticfoci described here is duplicated in axons exhibit-ing flattened vesicles .

The foregoing account has been primarily con-cerned with the organization of synaptic axo-plasm. Certain features of the cytoplasm of axonsmore or less remote from synaptic junctions mustbe noted . Irregular agranular cisternae, of variablewidth and orientation, occur in all Petromyzoncentral axons, but Bertolini's proposal (1964) thatthese are highly developed and connected with theplasma membrane has not been confirmed . Fur-thermore, no evidence has been obtained suggest-ing that synaptic vesicles are formed by fragmenta-tion of cisternae, microtubules or other componentsof the axoplasm . However, as illustrated in Fig . 14a-e, objects structurally identical with the sphericalvesicles concentrated at synapses occur with lowfrequency throughout the axoplasm, typically inassociation with microtubules. The axoplasm ad-joining the synaptic surface thus appears to bequantitatively rather than qualitatively distinctfrom other regions of the axon .

DISCUSSIONOne of the central problems in neurobiology con-cerns the functional relation between the peri-karyon and the often distant terminals of the nervecell . The complex cytoplasmic region surroundingthe nucleus maintains the cell, and net passage ofmaterials down the axon has been documented inearlier studies on the cellular behavior of neurons,as discussed by Weiss (1969) . However, it has be-come clear that some synthetic processes may occurat points outside the perikaryon, including thenerve terminals (see Barondes, 1969, for refer-ences). Recent studies have not only cast somelight upon the factors involved in directional move-ment of materials and structures within the cyto-plasm of a variety of cells but have also drawn at-

tention to oriented systems occurring in axons,other elongated cell processes, and elsewhere thathave been implicated as a possible framework fordirectional intracellular translocation .

While a wide variety of materials are un-doubtedly subject to oriented movement in cells,the process of translocation is most striking whenthe material moved is resolvable by light or elec-tron microscopy. The behavior of chromosomesupon the spindle is a long-established example ofprecisely directed intracellular translocation .Rebhun (1967) has described discontinuous orsaltatory non-Brownian movements of a variety ofgranules in invertebrate eggs and in melanocytesof Fundulus embryos, generally occurring at a rateof 0.5-2 Ä/sec . Bikle et al . (1966) further investi-gated the bidirectional movements of pigmentgranules in the melanocyte . Porter and Tilney(1965) have studied the translocation of particlesalong the slender axopodia of Actinosphaerium .Similar displacements of a variety of particles, in-cluding fat spheres and pinocytic droplets, havebeen described by Freed (1965) in culturedmammalian cells, and Holmes and Choppin (1968)have found that in virus-induced syncytia of kid-ney cells newly contributed nuclei migrate di-rectly to the center of the fused mass . The distalflow of axoplasm in nerve cells has been recog-nized for some time (Weiss et al ., 1962 ; Barondes,1969; Schmitt, 1968), and it now appears that inaddition to the massive but relatively slow bulkflow, specific materials may be transported farmore rapidly . The list of these materials alreadyreflects the importance of the cell body in main-taining and regulating the function of distantreaches of the cell, and includes rapidly trans-ported glutamate (Kerkut et al., 1967), proteins(McEwen and Grafstein, 1968) and also neuro-secretory granules synthesized in the vicinity ofthe nucleus (Jasinski et al ., 1966) .

The identification of a single structure under-

FIGURE 5 Micrograph illustrating the organization of axoplasm of a MUller cell (al) in the Petromyzonammocoete . The bulk of the axoplasm contains microtubules and neurofilaments (*) as in the smallaxons illustrated in Fig. 1, together with small mitochondria (m) . Peripherally, precisely demarcatedregions of the axoplasm are differentiated as synaptic foci, opposite small postsynaptic elements . Theseregions contain dense concentrations of synaptic vesicles. At 1 and $, vesicle clusters approach the axonplasma membrane very closely in regions circa 0 .5 Ä in width. These two synaptic foci are associatedwith a single postsynaptic (probably dendritic) nerve process (pl) while the clusters of vesicles at 3 and4 lie a short distance from the presynaptic surface and are evidently approaching points of synapse .At 5, a comparable group of vesicles is associated with a synapse between a small axon (a2) and a smallpostsynaptic process (p2) . Glial processes are indicated (g) . X 28,000 .

DAVID S. SMITH, U. JARLFORS, AND R . BERANEK Synaptic Axoplasm in Petromyzon

207

Dow


ber 2021

208

Dow


ber 2021

lying intracellular translocation is an attractive,but perhaps illusory goal . Electron microscopicstudies have shown that over a wide range of celltypes the most striking subcellular structures seenin thin sections in the region of transported cyto-plasm or particles are filaments and microtubules,respectively circa 50-100 and 200-250 A in di-ameter. While the latter have most frequently beenimplicated in transport mechanisms, there is evi-dence that filaments may sometimes be involved,and the picture is further complicated by the factthat each structure is composed of resolvable sub-units which have on occasion been considered asbuilding blocks common to both types of linearstructure .

In cytoplasmic droplets of the alga Nitella,Rebhun (1967) associated both saltatory andstreaming movements with bundles of 50 A micro-filaments . Bikle et al . (1966), on the other hand,observed microtubules extending along the armsof melanocytes ; they suggested that these do notact like the strings of a marionette, pushing andpulling the pigment granules, and moreover maynot merely act as a supporting framework for theasymmetric cell, but may act as guiding tracks formoving particles, and conceivably provide themeans of propelling these structures. They con-sider the microtubules as stationary structuresupon which the granules are moved and discussthe intriguing question of the operation of a single(central) or dual motive source in a two-direc-tional system. Porter and Tilney (1965) likewiseimplicated an elaborate arrangement of micro-tubules in the demarcation of regions of cytoplasmin the axopodia of Actinosphaerium within whichcytoplasmic movements occur. The nuclear migra-tion into a parainfluenza virus-induced syncytiumobserved by Holmes and Choppin (1968) has beencorrelated with microtubules and filaments, andisolated rows of nuclei have been obtained, ap-parently strung together by microtubules . Inouéand Sato (1967) have suggested that chromosomesmove against a superstructure provided by micro-tubules of the spindle, and Rudzinska (1965, 1967)

has proposed an elaborate microtubular array as aprobable pathway for translocation of digestiveenzymes and particles in the suctorian ciliateTokophrya .

If it could be established that microtubules andfilaments are self-sufficient, stable, and mutuallyexclusive structures, the question of their possibleinvolvement in intracellular transport would begreatly clarified . The structural uniformity ofmicrotubules is not paralleled by uniform stability .Behnke and Forer (1967) recognized four micro-tubule populations in insect spermatids, distin-guished by their response to conditions of fixation .Schultz and Case (1968) found that neuronalmicrotubules disintegrate in the presence of bi-carbonate ions and acrolein . The axopodia ofActinosphaerium are reversibly retracted by virtue ofdissociation of oriented microtubules by low tem-perature and high pressure (Tilney et al ., 1966 ;Tilney and Porter, 1967) . The disappearance ofthe 220 A microtubules in the cold is accompaniedby the formation of short segments of a largertubule 340 A in diameter, and at high pressure, bythe appearance of fibrillar material, supposedlyderived respectively by transformation and dis-integration of the normal structure . Holmes andChoppin (1968) described an increase in thecount of 80 A filaments accompanying the col-chicine-mediated disappearance of 250 A micro-tubules in virus-induced syncytia, and a similareffect occurs in interphase HeLa cells (Robbinsand Gonatas, 1964) . Microtubules and neurofila-ments appear to exist together, though in varyingproportions, in almost all axons throughout thephyla. There is a general reciprocity between pop-ulations of these components ; when one predomi-nates the other is relatively less conspicuous .Furthermore, as has already been mentioned, axonsize seems to play a part in determining the ratioof these two structures ; in the vertebrate, micro-tubules are generally most abundant in small un-myelinated axons, while in large myelinated axonsneurofilaments predominate . Schmitt and Samson(1969) consider the evidence suggesting that micro-

FIGURE 6 A synaptic region of Müller cell axoplasm (al) similar to that shown in Fig . 3, at highermagnification . This field includes a pair of synaptic foci involving two small postsynaptic elements(PI , p2) . The axon surfaces at the synapse are opposed across a cleft of 150 A (*) . Scattered microtubules(t) and neurofilaments (f) occur throughout the axon (cf. Fig. 3) except for a striking modification withinsynaptic vesicle clusters . Neurofilaments are sparse or absent within the synaptic foci, and microtubulesare restricted to the outer zone of vesicles, where they form the central element of eight rosettes (arrows),further illustrated in Fig . 7 . X 130,000 .

DAVID S. SMITH, U . JARLFORS, AND R . BERÂNEK Synaptic Axoplasm in Petromyzon

20 9

Dow


ber 2021

FIGURE 7 A micrograph further illustrating the association between synaptic vesicles and microtubulesin an axon of the Petromyzon central nerve cord . `Complete' rosettes comprise five synaptic vesicles and,in their immediate vicinity, lie randomly scattered vesicles and isolated microtubules. X 300,000 .

tubules and filaments may be interconvertible, gressively give rise to filaments-a view also putcomprising a common protein subunit . Peters and forward by Wisniewski et al . (1968) .Vaughn (1967) have supposed, from studies on

In assessing the possible role of longitudinallydeveloping neurons, that microtubules may pro- oriented structures in transporting materials along

FIGURE 8 A field within a longitudinally sectioned central nerve cord of a Petromyzon ammocoete,illustrating the striking alignment of synaptic vesicles within the axoplasm over a distance of about 5 ,u .Microtubules are evident within (short arrows) and alongside (long arrows) the vesicle groups . X 45,000 .

FIGURE 9 Further illustrating, in longitudinal aspect, the grouping of synaptic vesicles around micro-tubules in a Petromyzon axon. In the center of the field, vesicles are distributed in a close-packed hex-agonal pattern around microtubules, two or three of which lie within the plane of section . This specificassociation is revealed as the rosette pattern in transverse profiles (cf . Figs . 4, 5) . On the left lies anunaccompanied microtubule (t) and on the right, an elongated mitochondrion (m) . X 110,000 .

21 0

THE JOURNAL OF CELL BIOLOGY . VOLUME 46, 1970

Dow


ber 2021

211

Dow


ber 2021

FIGURE 10 Longitudinal section of an axon (al) of Petromyzon including a point of synapse with a post-synaptic nerve process (pi) . Synaptic vesicles (v) converge on the presynaptic membrane (*), but liefreely and apparently randomly in the axoplasm. Nearby, a single microtubule (t') is closely flanked by adouble row of synaptic vesicles (v')-an optical section of the close-packed configuration shown in Fig . 7 .X 130,000.

Dow


ber 2021

FIGURE 11 A micrograph further illustrating the distribution of synaptic vesicles in the vicinity of acentral synapse . The field includes an en passant synapse between a large axon (al) and small postsynapticcomponent (pi), across a synaptic gap indicated by an asterisk . Glial cytoplasm, containing numerousfilaments, is included at g. Abundant but unoriented synaptic vesicles (v) are clustered around the pre-synaptic focus at upper right. The left of the field is traversed by longitudinally aligned microtubulestightly ensheathed by synaptic vesicles (cf . Figs . 8-10) . In the center of the field, similar grouping occursabove the level of the synapse, but beneath, isolated microtubules (t') emerge . X 170,000.

Dow


ber 2021

214

Dow


ber 2021

the axon, it is important to establish whether ornot the obvious candidates, the microtubules andthe filaments, are chemically similar. Schmitt andSamson point out that neuronal microtubules aremore stable than structurally similar componentsof some other cells, being unaffected by exposureto low temperature (4 °C)-a finding confirmed inthe present Petromyzon material . On the other hand,neuronal microtubules are dissociated by colchi-cine or other mitotic inhibitors (Wisniewski et al .,1968) with the concomitant appearance of fila-mentous structures, supposedly abnormal poly-mers of the same protein aggregates normallyorganized into the microtubule . This does not,however, indicate that tubular and filamentouscomponents of normal axoplasm are interconverti-ble, since no rigorous method was employed fordistinguishing between a possible mixed popula-tion of normal and induced filaments.

There is, however, good reason to suppose thatneuronal microtubules (formerly termed "neuro-tubules") are homologous with microtubules ofother cells, including those implicated in intra-cellular movement, mentioned above . Borisy andTaylor (1967) found that colchicine-3H complexesspecifically with a protein in nervous (brain)tissue, cilia, and dividing cells, shown in the lastinstance (Shelanski and Taylor, 1967) to be thesubunit of microtubules. Weisenberg et al. (1968)have isolated from mammalian brain a proteinsimilar in a variety of biochemical properties tocilia protein, believed to be the building unit ofthe neuronal microtubule .

Schmitt and Samson (1969) stress that neuro-filament protein, investigated by Davison (1970)differs from that of microtubules (Borisy andTaylor, 1967) in lacking GTP-binding sites andcolchicine-binding properties characteristic ofmicrotubule subunits . Furthermore, Davison(1970) has recently provided the telling evidencethat squid axon neurofilament protein and micro-

tubule protein differ in their respective aminoacid composition, indicating that one componentcould only be derived from the other by syntheticreassembly of the constituent amino acids, and notby direct interconversion . On fine structuralgrounds, Wuerker (1970) strengthens the conclu-sion that the tubules and filaments of the axon arenot interconvertible structures, but rather twodistinct populations, each existing independentlyand presumably subserving separate functions .Wuerker and Palay investigated in detail the

distribution of neurofilaments and microtubules inrat anterior horn cells, with special reference todendrites . They confirmed that the microtubulewall contains 11-14 possibly inhomogeneous 60 Asubunits, leaving a central core of 150 A, some-times including a medial 50 A dot judged to be afilamentous structure traversing the microtubule(cf. Echandia et al ., 1968). The neurofilaments,on the other hand, are reported to be circa 100 Ain diameter, likewise including a translucent core,but comprising a wall of three to six subunits, eachonly 30 A in diameter . Wuerker and Palay pointout that direct transformation of one componentinto the other (as by unravelling of the singletubule to provide neurofilaments) cannot beachieved by any obvious sequence of events . Theyalso point out, in support of the nonequivalence ofmicrotubules and filaments, that depolymerizationof tubules by low temperature (Echandia andPiezzi, 1968) produces a fibrous material distinctfrom the neurofilaments, which persist unaffectedby the experimental treatment . Wuerker andPalay are especially concerned with the possiblerole of neurofilaments as transporting organelles .They have correlated the observations of Pomeratet al . (1967) on bidirectional streaming within thecell body with channels occupied by neurofila-ments. In certain neurons, Wuerker and Palayfurther report a close association between bundlesof neurofilaments and Nissl bodies and suggest that

FIGURE 12 An axon within the spinal cord of Petromyzon which contains, in addition to the sphericalelectron-lucent synaptic vesicles illustrated in Figs . 3-11, a second axoplasmic component, indicatedby arrows. This takes the form of a larger membrane-limited structure, circa 1000 A in diameter, in-cluding a dense core of variable size, partially filling the vesicle . A mitochondrion is included at m .X 130,000.

FIGURE 13 A field similar to the last, but including a point of pre- and postsynaptic membraneapposition (*) . Spherical lucent vesicles are primarily involved in the synaptic focus, but occasionallarger structures (arrow) (cf . Fig . 11) containing an irregular core of dense material are included, butdo not seem to be preferentially located near the synaptic axon surface . X 120,000.

DAVID S . SMITH, II. JXRLFORS, AND R. BERe1NEK Synaptic Axoplasm in Petromyzon

21 5

Dow


ber 2021

the production of the bundles may be controlledby the Nissl cisternae, and that neurofilamentsmay be actively concerned in the centrifugal trans-port of materials. They have described the micro-tubules, in mammalian material, as arranged atintervals of 100 it and suspended in a web oflateral threads . In the lamprey material, lateralthreads link the neurofilaments, but neither thissuspensory web nor comparable regularity marksthe microtubules .

It is entirely possible that the neurofilamentsperform a transporting function in all nerve cells,following the model proposed by Wuerker andPalay. The morphological evidence they presentis impressive and the neurofilament is certainly aconspicuous and virtually universal component ofneurons. For several reasons, these authors con-clude that microtubules and neurofilaments aredistinct, noninterconvertible entities . Schmitt andSamson (1969) have considered biochemical evi-dence that leads to the same conclusion . Fromobservations on the ammocoete nerve cord ma-terial, we feel that while the generalization con-

21 6

FIGURE 14 a-e A group of electron micrographs representing fields within large axons in the Petromyzonspinal cord . These areas do not adjoin synapses . Close examination of the nonsynaptic axoplasm revealselectron-lucent vesicles, singly or in small groups, preferentially located beside microtubules . As illus-trated here, these are structurally similar to the spherical synaptic vesicles included in Figs . 5-11 . In axonspossessing ellipsoidal or discoidal vesicles in the synaptic foci (see text), similar vesicles are sparselyassociated with microtubules outside the synaptic areas . In both of these neuron types, larger vesicleswith an electron-opaque core (Figs . 12, 13) have occasionally been noted beside microtubules in areassome distance from synaptic terminals . X 80,000 .

THE JOURNAL OF CELL BIOLOGY • VOLUME 46, 1970

cerning proportionality of filaments and micro-tubules vis-â-vis size of nerve profile holds good ;the ratios, reaching a limiting count of 20-30 :1 donot show any evidence of direct or simple inter-conversion such as dissociation of a microtubuleinto filamentous subunits .

We are conditioned to movement along restrain-ing pathways-turnpikes, cycle tracks, escalators,footpaths, and so on, each providing for an appro-priate speed range or type of traffic . It may wellbe that intracellular traffic in the neuron is analo-gously varied . This cell type contains two primecandidates for providing intracellular tracks, themicrotubules and neurofilaments, and while theevidence suggests that these are self-sufficient anddistinct structures, it should also be rememberedthat both fast and slow transport occurs concur-rently and selectively in the axon. It is temptingto suppose that not only do the axoplasmic fibrousproteins of the axoplasm play a part in transport,but that the filaments and tubules may representdual transport pathways . Schmitt (1968) has sug-gested that the slow, nonspecific bulk flow of axo-

Dow


ber 2021

plasm away from the cell body may be effected by

the neurofilaments, and that rapid and specifictranslocation, in neurons and other cells, may becontrolled by microtubules . Schmitt (1968) andSchmitt and Samson (1969) further proposed thatmicrotubules extend as stationary tracks from cellbody to extremities of the neuron, and that theymay incorporate part of a mechanochemical trans-duction mechanism responsible for rapid trans-port. According to their hypothesis, a microtubule-particulate association functionally and perhapsbiochemically comparable with myosin-actininteraction of myofilaments may be responsible formoving certain components along the axon at ratesup to several hundred times that of the slowlymoving axoplasm, and they discuss the possibilitythat directionality is imposed by the sense of thehelical disposition of subunits of the axonal fibrous

systems .

The results described here in no way conflictwith the above suggestions, and they document,for the first time, an association between micro-tubules and another axoplasmic structure ; theseresults are remarkably close to diagrammatic rep-resentations outlined by Schmitt and Samson. Tointerpret the accompanying electron micrographs,some assumption must be made and recognized .(a) Preliminary counts indicate that microtubuledistribution is similar in synaptic foci to that inthe general axoplasm, even in the "giant" neuronsof Petromyzon . We have no reason to suppose thatmicrotubules are divided into two populations, onedestined to reach synapses and another not, so weassume that each synapse receives its due comple-ment of microtubules, presumably stemming fromthe perikaryon . Thus a microtubule in the centerof, for example, a giant Müller fiber in the am-mocoete is envisaged as being en route for one orprobably more synapses further along the spinalcord . (b) Current observations on the Petromyzoncord indicate that microtubules do not terminateat synapses ; electron micrographs repeatedly showthese tubules bypassing the presynaptic membraneat the periphery of the synaptic focus . This is mostobvious in the giant fibers, but occurs also insmaller units . A giant fiber 20,u in diameter (suchas that illustrated in part of Fig. 5) contains ap-proximately 1500-3000 microtubules counted in

transverse profile (together with some 75,000-

100,000 neurofilaments) . Individual synaptic foci

commonly include about five microtubules . If thetubules ended at the foci, such a giant cell might

contain of the order of 300-600 synaptic junctionswith small nerve processes (primarily dendrites),

but the arrangement of microtubules appears toensure that each is associated with a sequence offoci, and the count of synaptic junctions within acell is unknown and probably very much greaterthan the above figure . (c) Electron-lucent vesicles,

whether spherical or discoidal, occur sparselythroughout the axoplasm, but become regularlydisposed around the microtubules in the vicinityof synapses. This observation is consistent withthe hypothesis that the vesicles are passed alongthe tubules to accumulate opposite locally differ-entiated regions where release of chemical trans-mitters takes place. Since the tubules do not ap-pear to be confined to a single synaptic region, it ispossible that each region may be supplied withvesicles only to a hypothetical saturation point,permitting overspill and sequential distribution of

vesicles to more than one synapse . (d) If we assumethat the association of vesicles with microtubulesreflects a mechanism for vesicle transport, then thefocal clustering of "free" vesicles immediately ad-joining the presynaptic surface may indicate thatvesicles are released from the adjacent micro-tubules, perhaps destined for nearby receptive siteson the axon membrane (postulated by Katz, 1962 ;Hubbard et al ., 1968) at which transmitter releasetakes place . (e) The speculative nature of these in-terpretations is clear, and it should be pointed outthat the chemical nature of the transmitters presentin the lamprey spinal cord are not yet known . Weare as yet unable to assess, in this material, therelative importance of the perikaryon and exten-sions of the neurons, including synaptic regions, intransmitter synthesis .The presence of 700-1200 A vesicles with a

dense core in lamprey neurons deserves comment .Smaller dense-cored vesicles are established as aconspicuous component of adrenergic nerves, butdistinct larger-cored structures have been re-ported in a variety of neurons not generallyrecognized as hormone secreting . De Iraldi andDe Robertis (1968) propose that these are prob-ably present in small numbers in all types ofneuron but found them to be particularly abun-dant in nerves regenerating after damage, a find-ing confirmed by Van Arsdall and Lentz (1968) .

Richardson (1964) also noted that the distribution

of these cored vesicles may include cholinergicneurons. Their significance in the lamprey spinal

cord and elsewhere, and in particular their possi-

DAVID S. SA'IITH, U. JÄRLFORS, AND R. BERANEK Synaptic Axoplasm in Petromyzon

217

Dow


ber 2021

ble relationship to the smaller lucent vesicles andto the synapse, has yet to be established .

Links between cellular structure and functiongenerally pay little regard to phyletic boundaries .Certainly, it is expected that so fundamental afunction as rapid and specific transport in nervecells of diverse animals would share a commonstructural basis . It will be important to ascertain,in the future, whether the relationship betweenmicrotubules, vesicular components of the axo-plasm, and presynaptic areas, described in Petro-myzon, occurs in the nervous system of other animalsor indeed throughout the larval and adult life ofthe lamprey. Meanwhile, the ammocoete promisesto afford convenient material for continued experi-mental studies on the organization and functionof microtubules and other intracellular compo-nents of nerve cells.

RADAN BERANEK . 1923-1969 . Dr. Berânek washead of the Laboratory of Cellular and ComparativeNeurophysiology, the Czechoslovak Academy ofScience, Prague . He came to the University of

REFERENCES

ATWOOD, H. L., and A . JONES. 1967 . Experientia(Basel) . 23 :1036 .

BARONDES, S . H. 1969. Neurosciences ResearchSymposium Summaries . F. O. Schmitt, T . Mel-nechuk, G . C. Quarton, and G . Adelman, editors.The M.I .T. Press, Cambridge, Mass. 3:191 .

BEHNKE, O., and A . FORER . 1967 . J. Cell Sci. 2 :169.BERTOLINI, B. 1964 . J. Ultrastruct . Res. 11 :1 .BIKLE, D ., L. G. TILNEY, and K. R. PORTER . 1966 .

Protoplasm. 61:322 .BORISY, G. G., and E. W. TAYLOR. 1967. J. Cell

Biol . 34 :525 .DAVISON, P. S. 1970. Microtubules and neurofila-

ments: possible implication in axoplasmic trans-port. In Biochemistry of Simple Neuronal Models .Raven Press, Hewlett, N . Y. In press .

DE IRALDI, A. P., and E . DE ROBERTIS . 1968. Z.Zellforsch . Mikrosk . Anat . 87:330 .

DE ROBERTIS, E. 1958. Exp. Cell Res . 5 :(Suppl .) 347.DE RoBERTIS, E. 1967 . Science (Washington) . 156:907.ECHANDIA, E. L. R., and R, S . PIEzzi . 1968 . J. Cell

Biol . 39 :491 .ECHANDIA, E . L. R., R . S. PIEZZI, and E. M. RoD-

RIGUEZ . 1968. Amer. J. Anat . 122 :157 .FREED, J . J . 1965. J. Cell Biol . 27 :29A .GONATAS, N . K., and E . ROBBINS. 1965 . Protoplasma .

59:377.HOLMES, K. V., and P. W. CHOPPIN . 1968 . J. Cell

Biol. 39 :526 .

2 1 8 THE JOURNAL OF CELL BIOLOGY . VOLUME 46, 1970

Miami School of Medicine in November, 1968,as visiting Professor of Physiology, at which time heintroduced us to ammocoetes and we started afine structural study of the central nervous system ofPetromyzon . We were able to discuss several of thefeatures described in this paper and had begun toconsider longer term studies at the time of RadanBerânek's death in March 1969. He, and his collabo-ration, is greatly missed .

This work was supported by grant GB-12117Xfrom the National Science Foundation .

Received for publication 10 December 1969, and in revisedform 26 January 1970 .

Note Added in Proof : Electron microscopic evidencehas recently been obtained of structural links betweensynaptic vesicles and microtubules in the ammocoetenerve cord. These preliminary findings, made afterthe present paper went to press, will be reported ata Discussion Meeting of the Royal Society of Londonon "Subcellular and macromolecular aspects of syn-aptic transmission" held in London, May, 1970. (D .S. Smith, Proc . Roy . Soc. (London) Ser . B ., manuscriptin preparation) .

HUBBARD, J. I ., S . F. JONES, and E . M. LANDAU .1968 . J. Physiol. 196 :75 .

INQUk, S ., and H . SATO . 1967. J . Gen . Physiol . 50:259.JARLFORS, U ., and D . S . SMITH . 1969 . Nature (London) .

224 :710 .JASINSKI, A ., A. GORBMAN, and T. J. HARA. 1966.

Science (Washington) . 154 :776 .KATZ, B. 1962 . Proc . Roy . Soc . (London), Ser. B . 155 :455 .KERRUT, G . A., A. SHAPIRA, and R . J. WALKER .

1967 . Comp. Biochem . Physiol. 23 :729 .LANE, N . J ., and J . E. TREHERNE . 1969. J. Cell Biol .

43 :76a .MCEWEN, B. S ., and B. GRAFSTEIN . 1968 . J. Cell

Biol . 38 :494 .NAKAJIMA, Y . 1970. Tissue Cell. 2 . In press .PALAY, S. L. 1958 . Exp. Cell Res . 5(Suppl .) :275 .PALAY, S . L. 1964 . The structural basis for neural

action. In RNA and Brain Function : Memory andLearning (Brain Function) . M. A. B, Brazier,editor . University of California Press, Berkeley .2 :69 .

PETERS, A ., and J . E. VAUGHN. 1967 . J. Cell Biol .32 :113 .

POMERAT, C. M., W . J . HENDELMAN, C. W. RAIBORN,and J . F. MASSEY . 1967 . Dynamic activities of thenervous tissue in vitro . In The Neuron. H. Hydén,editor . Elsevier, Amsterdam . 119.

PORTER, K. R., and L . G. TILNEY . 1965 . Science(Washington) . 150 :382 .

Dow


ber 2021

REBHUN, L. I . 1967 . J. Gen . Physiol . 50 :223 .RETZivs, G. 1891 . Biol. Untersuch . Stockh . 2 (n . s .) :47.RICHARDSON, K. C. 1964 . Amer . J. Anat. 114 :173 .ROBBINS, E ., and N . K . GONATAS . 1964 . J. Histochem .

Cytochem . 21 :429.ROVAINEN, C . M . 1967. J. Neurophysiol . 30:1000 .RUDZINSKA, M . A . 1965 . J. Cell Biol . 25 :459 .RUDZINSKA, M . A . 1967 . Ann . N. Y . Acad. Sci . 29 :512 .SCHMITT, F. 0. 1968 . Proc . Nat . Acad. Sci. U.S.A .,

60:1092.SCHMITT, F. 0 ., and F. E. SAMSON . 1969 . Neuro-

sciences Research Symposium Summaries. TheM.I.T. Press, Cambridge, Mass. 3 :301 .

SCHULTZ, R . L., E . C . BERKOWITZ, and D. C . PEASE .1956 . J . Morphol. 98 :251 .

SCHULTZ, R . L ., and N. M. CASE . 1968. J. CellBiol . 38 :633 .

SHELANSKI, M . L ., and E. W. TAYLOR . 1967. J .Cell Biol. 34:549.

TILNEY, L. G., Y. HIRAMOTO, and D . MARSLAND .1966. J. Cell Biol . 29 :77.

TILNEY, L. G ., and K. R. PORTER . 1967. J. CellBiol . 34 :327 .

UCHIZONO, K. 1965. Nature (London) . 207:642 .UCHIZONo, K . 1967. Nature (London) . 214:833 .VAN ARSDALL, C. B., and T. L . LENTZ . 1968 . Science

(Washington) . 162:1296.WEISENBERG, R . C., G . G . BORISY, and E. W. TAYLOR .

1968 . Biochemistry . 7 :4466 .WEISS, P ., A . C . TAYLOR, and P . A . PILLAI . 1962 .

Science (Washington) . 136 :330.WEISS, P. 1969 . Neurosciences Research SymposiumSummaries, F. 0. Schmitt, T. Melnechuk, G. C.Quarton, and G . Adelman, editors. The M .I .T.Press, Cambridge, Mass . 3 :255 .

WHITTAKER, V. P. 1968 . Proc . Nat . Acad. Sci. U.S.A .60:1081 .

WIGGLESWORTH, V. B . 1960. J. Exp . Biol. 37:500.WISNIEWSKI, H., M . L . SHELANSKI, and R . D . TERRY.

1968 . J . Cell Biol. 38:224.WUERKER, R . B., and S. L . PALAY . 1969. Tissue

Cell . 1 :387 .WUERKER, R . B. 1970. Tissue Cell. 2 . In press .

DAVID S. SMITH, U. JXRLFORS, AND R. BER1NEK Synaptic Axoplasm in Petromyzon

219

Dow


ber 2021

the organization of synaptic axoplasm in the lamprey … · the filial filaments described by...

Documents