Initiation of fast axonal transport: Involvement of calcium during transfer of proteins from golgi apparatus to the transport system

INITIATION OF FAST AXONAL TRANSPORT: INVOLVEMENT OF CALCIUM DURING TRANSFER OF PROTEINS FROM GOLGI APPARATUS TO THE TRANSPORT SYSTEM R. HAMMER~CHLAGand P.-A. LAVOIE’ Division of Neurosciences, City of Hope National Medical Center, Duarte. CA 91010, U.S.A. Abstract-Fast axonal transport of [‘Hlfucose-labelled glycoproteins was examined in uirro in a preparation of spinal sensory neurons of the bullfrog. Rapidly transported glycoproteins were labelled when dorsal root ganglia were exposed to [3H]fucose, but not when a localized region of spinal nerve trunk was exposed to the labelled sugar; these results are consistent with the general finding that fucosylation occurs predominantly in the Golgi apparatus. Incubation of ganglia and nerve trunks in medium containit& 0.18 mM CoCl, depressed the amount of rapidly transported. [3H]fucose-labelled glycoptotein by approximately 80%. This level of cobalt impairs neither protein synthesis nor incorporation of [“H]fucose into glycoprotein. suggesting that the inhibition of axonal transport by cobalt occurs at a step subsequent to fucosylation. When a 4mm region of spinal nerve was desheathed, to facilitate access of ions to the axons. cobalt had no effect on fast axonal transport of glycoproteins at the level of the axon: the amount of [3H]fucose-labelled protein was similar in the regions of nerve trunk proximal and distal to the desheathed area. Thus, the cobalt-sensitive step in the transport of glycoproteins is likely to occur in the somata during transfer of the proteins from the Golgi apparatus to the transport system: it is suggested that cobalt antagonizes the action of intracellular calcium ions. An analogy is considered between the processing of fast-transported proteins by neuronal somata and the processing of secretory proteins by pancreatic exocrine cells.

WHILE most

studies on fast axonal transport of protein have been concerned with the translocation process within the axon (for reviews see OCHS 1974; HESLOP, 1975; GRAFSTEIN, 19773, relatively

little attention

has been focused on how transport is initiated in the soma. This initial phase of transport comprises the sequence of events by which proteins destined for fast transport migrate from their sites of synthesis to the transport system. Based on electron-microscopic radioautography. it has heen proposed that polypeptide chains synthesized in the somal rough endoplasmic reticulum, including proteins that will undergo fast axonal transport, initially migrate to the Golgi apparatus where glycosylation of selected proteins is completed (DROZ. 1965; 1975; DROZ & KOENIG, 1970). These ultrasttuctural studies have not specified how the proteins destined for axonal transport could be transferred from the Golgi apparatus to the transport system. The present paper offers biochemical ev’lence that this transfer is a calcium-dependent, cobalt-sensitive process. Previous studies have demonstrated that the requirement for calcium during the initiation of axonal transport occurs after protein synthesis (DRAVID & HAMMERSCHLAG.1975a; HAMMERSCHLAG,DRAVID & CHIU, 1975). A

next step was to determine

whether

’ Present address: Dkpartement de Pharmacologic. Universitk de Mont&al, Case Postale 6128 Succursalc ‘A’. Montreal. Canada H3C 337.

calcium is required before or after arrival of the proteins at the Golgi apparatus. This was studied by following the effects of cobalt, an antagonist of calcium in axonal transport and other processes (HAMMEASCHLAG, CHIU & DRAVID. 1976), on the incorporation and subsequent axonal transport of [3H]fucose. Such an approach seemed justified because (i) the calciumdependent step during initiation appears common to all rapidly transported proteins (DRAVID & HAMMERSCHLAG. 197563, many of which are glycoproteins (KARLSSON& SJ&~RAND, 1971; EDSTR~~M8 MAITSSON, 1973; BARKER, HOFFMAN,GAINER & LASEK. 1975); and (ii) fucose is located in a terminal position of carbohydrate side chains (SPIRO, 1970), and terminal glycosylation occurs in the Golgi apparatus (BARONDES,1970; LEBLOND& BENNETT,1977). Previous studies have also suggested that the calcium requirement occurs prior to the onset of the translocation phase of transport (HAMMERSCHLAG er al., 1975; HAMMERSCHLAG, BAKHIT,CHIU & DRAVID, 1977). While this conclusion was drawn from studies on intact nerves, calcium-free incubation conditions have recently been found to block fast axonal transport in a desheathed mammalian nerve (0~~s. WOR-ITI & CHAN, 1977). Since desheathing facilitates access of ions to the axons (KRNJEVI~, 1954; FRANKENHAEUSER,1957), it was of interest to test whether such a procedure could reveal an inhibition by cobalt ions similar to their effect on ganglia (HAMMERSCHUG et al.. 1977). If cobalt does depress transport through a desheathed retion of nerve. it could be inferred that

1195

1196

R. HAMMERSCHLAG and P.-A. LAVOIE

the action of cobalt on the ganglia is on the proximal portion of axons rather than in the somata.

small ditferences in the total incorporation of [‘Hlfucose between control and experimental preparations.

EXPERIMENTAL PROCEDURES General experimental procedure. The 8th and 9th dorsal

RESULTS Site of incorporation

of fucose

into rapidly

transported

root ganglia of the bullfrog, Rana catesbeiana. were disproteins sected in continuity with their respective spinal roots Accumulation of acid-insoluble radioactivity was spinal nerves and sciatic nerve, and placed in a multi-comassessed in ligated nerve trunks after 1 h pulse-labeipartment Lucite chamber (DRAMD & HAMMJZRSCHLAG, 1975a). Dorsal root ganglia or spinal nerves were selec- ling of ganglia with [ ‘Hlfucose and 17 h of subsequent tively exposed to [3H]fucose (6OCi/mmol; New England incubation in normal medium (Fig. 1A). Build-up

Nuclear Corp.) contained in normal medium of the following composition (mM): NaCl, 114; KCl, 2.0; CaCI,, 1.8; glucose, 5.5; 4-(2-hydroxyethylbl-piperazine-ethanesulfonic acid (pH 7.4). 20.0. Final concentration of [‘HJfucose was either 2.08 PM or 8.32 C(M;results with both concentrations were qualitatively similar. After pulse-labelling and sub sequent incubation at 18°C in oxygenated medium, nerves were cut into either 2mm (ligated nerves, see below) or 3 mm (unligated nerves) segments, and the trichloroacetic acid-insoluble radioactivity of each ganglion and nerve segment was determined as previously described (DRAMD& HAMMERSCHLAG, 1975a). Accumulated radioactivity proximal to a ligature or total acid-insoluble radioactivity in nerve trunks were the usual indices of transport. Selective exposure of spinal nerves to [‘Hlfiose. In experiments to test for incorporation of C3H]fucose into fasttransported glycoproteins in axons. ganglia were removed to prevent any incorporation into the somata. A ligature was placed on the sciatic nerve 12 mm distal to the junction of the 8th and 9th spinal nerves. The region of the spinal nerves 1619mm proximal to the ligature was exposed to C3H]fucose for 3-5 h at 22°C. Fast-transported proteins (potential fucose acceptors) passed through this region of nerve trunk during the entire pulse-labelling period. The isolated nerve trunks were then incubated in normal medium at 18°C for 18 h during which time the preparation was continuously superfused with oxygenated normal medium. Exposure of intact or locally desheathed preparations to cobalt-containing medium. In intact preparations, ligatures

were tied on each 8th spinal nerve at 18 mm from the ganglion, while the 9th nerves were left unligated so that transport could be followed along the sciatic nerve. Ganglia were pulse-labelled with [‘Hlfucose for 1 h at 22°C. One preparation was then transferred to normal medium containing 0.18m~ CoCI,, while the contralateral prep aration was placed as a control in normal medium. Incubations were carried out for 1I h at 18°C with oxygenation. In other experiments, spinal nerves were locally desheathed to facilitate entry of cobalt ions into axons. Desheathing (FENG & LIIJ. 1949) was performed on preparations pinned out via ligatures on terminal regions of spinal roots, and at 28 mm from the ganglion on the spinal nerve. The nerve was desheathed between 16 and 20mm from its ganglion. Following a 1 h period of exposure of ganglia to [‘H]fucose, preparations were incubated for 17 h at 18°C in normal medium (control) or in medium containing either 0.18 or 0.018 mM CoClx (experimental). Effects of cobalt on transport of [‘Hlglycoproteins were then compared in the region proximal to the desheathed area, and at the ligature distal to the desheathed region of the nerve. In the calculation of percentage inhibition, radioactivity in control nerve segments was corrected for

proximal to the ligature was readily detected, a finding consistent with the fast transport of [3H]fucoselabelled glycoproteins previously reported (ZATZ & BARONDES, 1971; KARLSSON & WSTRAND, 1971; FORMAN, GRAFSTEIN& MCEWEN, 1972; EDSTROM& MATSWN, 1972; MARKO& CUJ~NOD, 1973; BENNETT, GIAMBERARDINO,KOENIG& DROZ, 1973 ; LANGLEY&

CARMICHAEL. 1978). Since local incorporation of terminal carbohydrate residues has been described in Aplysia axons (AMBRON& TREISTMAN,1977), it was of interest to assess whether any appreciable C3H]fucase might be added to fast-transported protein in the proximal region of axons within bullfrog ganglia. On the assumption that the region of axons in nerve trunks can serve as a model for the intraganglionic region of axons, a 3 mm region of isolated nerve trunk was selectively exposed to [3H]fucose for 3 h followed by an 18 h incubation period. While substantial local incorporation of the isotope was observed, no accumulation of [3H]glycoproteins was detected at a ligature distal to the site of pulse-labelling (Fig 1B). This local incorporation is presumably into glial proteins and, conceivably, into non-transported axonal proteins. Since there is no evidence for any appreciable local incorporation of fucose into glycoproteins undergoing fast transport, the site of fucosylation is likely to be the Golgi apparatus, in keeping with observations in numerous other systems (LEBLOND& BENNEIT,1977). Eficts

of cobalt

chloride

on transport

and synthesis

of glycoproteins The amount of C3H]fucose-labelled glycoproteins undergoing fast axonal transport during 18 h was depressed by approximately 80% when intact preparations were incubated in medium containing 0.18m~ CoCI, (Table 1). Transport was decreased to a similar extent in both ligated and unligated preparations. Radioactivity profiles of unligated nerves also revealed that the reduced amount of C3H]glycoproteins transported in nerves incubated in cobaltcontaining medium travelled at a normal rate, as previously observed for C3H]leucine-labeiled proteins (HAMMERSCHLAG et al., 1976). The decreased amount of [3H]glycoproteins exported into the axons during incubation with cobalt was balanced by an increased level of acid-insoluble radioactivity in the ganglia (Table 1).

1197

Initiation of fast axonal transport

; \ .

‘\.\ 1 .

\ .

lL_.;

i

-

I

10

10 DISTANCE GANGLION

+ 210

3’0

FROM (mm)

b

20

DISTANCE

10 FROM

LIGATURE

3

z (mm)

FIG. 1. Axonal transport of [3H]glycoproteins after exposure of ganglion (A) or nerve trunks (B) to [‘H]fucose. In (A). the ganglion was pulse-labelled for 1 h, and the preparation incubated for a further 17 h, while in (BL a 3 mm region (indicated by the double-headed arrow) of isolated 8th and 9th spinal nerves was simultaneously pulse-labelled for 3 h and the preparation incubated for an additional 18 h. Each point represents trichloroacetic acid (TCA)-insoluble radioactivity present in a ganglion or 2 mm segment of nerve. In (B), points represent 8th spinal nerve (M). 9th spinal nerve (o--O), or sciatic nerve (A-A). The vertical arrows indicate the position of the ligature on the nerve trunk. The examples shown are representive of six experiments for (A) and five for (B).

The extent of [)H]fucose incorporation into glycoproteins after 1 h and after 18 h is shown in Table 2. Close to 90% of the total incorporation occurred during the 17 h incubation period. The amount of trichloroacetic acid-soluble radioactivity in ganglia after the 1 h pulse-labelling period was ample to account

for the extent of this subsequent incorporation into glycoproteins. Such a prolonged period of incorporation of [3H]fucose has been postulated to explain an extended time-course of arrival of glycoproteins at nerve terminals (KARLSSON& WSTRAND, 1971; FORMANet al., 1972). Total trichloroacetic acid-inso-

TABLE 1. EFFECT OF INCUBATION OF INTACT PREPARATIONSIN MEDIUM CONTAINING 0.18 mM CoClz ON SYNTHESIS AND TRANSPORT OF [3H]~~c~~~-~~~~~ GLYCOPROTEINS

Preparation 8th spinal ganglion and spinal nerve (ligated) 9th spinal ganglion, spinal nerve (unligated). and sciatic nerve

c.p.m. experimental preparation cXm. control nreoaration Nerve ‘trunks Ganglia Total 1.17 + 0.10*

0.25 f 0.04

1.00 * 0.10

(6) 1.14 + 0.05

(6) 0.17 k 0.02

0.96 k 0.04

(7)

(7)

* Values represent mean + S.E.M.for preparations pulse-labelled for 1 h with [3H]fucose and incubated for 17 h in experimental or control medium; the number of experiments is shown in parentheses. c.p.m.. counts per minute.

R. HAMMERSCHLAGand P.-i\. LAVW

1198

TABLE 2. UPTAKE AND INCorwoBAnON OF [3H]~~~~~~ INTO GLYCOPROTEINAFTER 1h OF PULSE-LABELLING, AND AFTER A SUBSEQUENT 17 h OF INCUBATION

Time (h)

I 18

TCA-insoluble

TCA-soluble

radioactivity

radioactivity

3.410 &- 367 (S)* 26,001 & 2,291(6n

79,668 & 10,996 (8)* Not determined

* c.p.m./ganghon: mean + S.E.M.;number of samples in parentheses. Following the I h pulse-labelling period, ganglia were separated from

their spinal roots

and spinal

nerves. rinsed in isotope-free normal medium for 15 s, and transferred to cold 5% TCA. Samples were processed as previously described (DRAVID & HAMMERSCHLAC,1975a). TCA, trichloroacetic acid. t c.p.m.ganglion plus nerve segments: mean + S.E.M.(n).

BOLEN & HAMMERSCHLAG. 1979). controlnervesin the present experiments were also desheathed. The level of radioactivity proximal to the desheathed area was depressed by approximately 854, due to an action of 0.18 mM CoCl, on the ganglion (Table 3A). A similar level of inhibition by cobalt was observed at the ligature. where accumulation of radioactivity depends on the passage of labelled glycoproteins through the desheathed region. Since the depression of transport at this concentration of Co’+ was not maximal (1.8 mM CoC12 causes virtually complete blockage of transport [LAVOIE. HAMMERSCHLAG & TJAN, 19781). an additional effect of Co” on the desheathed area would have been detected at the ligature. When 0.018 I'tIM CoCI, was used (to produce a less drastic inhibition in the ganglion) transport again was depressed to a similar extent in the regions of nerve trunk proximal and distal to the desheathed area (Table 3A). Thus. since no additional depression of transport was detected at the ligature, there was no indication that CoL’ depresses fast transport at the level of the axon. This observation was confirmed in experiments where, following the pulse-labelhng period. the locally desheathed nerve trunk was exposed to medium containing 0. I8 rnM CoCl, while the ganglion was maintained in normal medium (Table 3B). No depression of transport was observed proximal or distal to the desheathed region of nerve

luble radioactivity of preparations incubated for 18 h in normal medium was similar to that of preparations exposed to cobalt-containing medium (Table l), indicating that cobalt has no effect on the process of fuco sylation. While it is conceivable that cobalt could selectively depress fucosylation of proteins destined for fast transport, such an effect would probably have been detected since acid-insoluble radioactivity transported in 18 h accounts for 18 f 2% (mean & s.E.M.; II= 12) of total incorporated C3H]fucose. To determine whether cobalt can depress the fast transport of [3H]glycoproteins in the axon, preptrunk. arations in which each spinal nerve was locally desheathed were incubated in cobalt-containing medium. Since desheathing per se causes a slight imThe intracellular pairment of transport (OCHS et al., 1977; LAVOIE,

DISCUSSION

sequence of events by which pro-

TABLE 3. EFFECT OF INCUBATIONIN COBALT-CONTAKING MEDIUM ON FAST AXONAL TRANSPORT OF ['H]FUCOSE-LABELLED GLYCOPROTEINS IN REGIONS OF AXONS PROXIMAL AND DISTAL TO A DESHEATHED REGION OF NERVE TRLINK

Portion of preparation exposed to cobalt A.

Ganglion and

Cobalt ion concentration (mW

c.p.m. experimental preparation c.p.m. control preparation Pre-desheathed Ligature region region

nerve trunk*

0.18 0.018

0.15 + 0.04 0.46 -+ 0.1 I

0. I7 + 0.02 (4) 0.48 + 0.09 (3)

B. Nerve trunk*

0.18

1.08 + 0.10

1.01 + 0.06(9)

* In A: After 1h pulse-labeiling the experimental preparation was transferred to cobalt-containing medium, and the control preparation was transferred to normal medium. In B: After 1 h of pulse-labelling. the experimental preparation was transferred to a Lucite chamber that permitted selective exposure of the nerve trunk to cobalt-containing medium. while the ganglion was maintained in normal medium. In both sets of studies. incubatmns were carried out for an additional 17 h, at which time nerve trunks were cut into 2 mm segments starting from the ligature at 28 mm. Levels of trichloroacetic acid-insoluble radioactivity in the 2 mm nerve segment proximal to the ligature (L). and in segments from the region IO-14mm from the ganglion. were compared for paired preparations incubated in normal medium or in cobalt-containing medium after normalization of c.p.m. as described in Experimental Procedures. The desheathed region (double-headed arrow) was 16-20 mm from the ganglion (G). Values are mean + S.E.M.; number of experiments in parentheses.

Initiation 0r

fast axonal transport

teins destined for fast axonal transport are transferred from their sites of synthesis to the transport system is not well understood. Such events in the ceil body can be studied in the spinal sensory nerve preparation even though the proximal regions of axons also he within the ganglion. Our approach has been to compare effects of transport in nerve trunks and in the ganglia, and to assume that events within the region of axons in the nerve trunk are similar to those in that portion of axons within the ganglion. Accordingly, since fast transport is unaffected by the incubation of intact nerve trunks in calcium-free medium or cobalt-containing medium, it was proposed that the impatient of transport following exposure of ganglia to these media occurs at the cell body (HAMM~~~SCHLAG CT al..1975; 19771. This interpretation was questioned by the finding that calcium-free incubation conditions will block axonal transport of C3H]proteins in desheathed peroneal nerves of cat (OCHS et al., 19771. Since, in the present studies, cobalt was not found to depress transport at the level of the axons. it seems justified to infer that the site of action of cobalt within the ganglion is in the somata rather than in the intra~~glioni~ portions of axons. To define this site further, the effects of cobalt on synthesis and on fast axonal transport of [3H]fucoselabelled glycoproteins were compared. Incubation of ganglia in cobalt-containing medium markedly reduced the fast transport of [‘H]glycoproteins. While incorporation of C3H]fucose took place predominantly during this incubation period, the inhibition of transport on exposure to cobalt occurred with no concomitant depression of fucosy~tion. This dis” sociation of effects on synthesis and transport, which was also observed when a mixture of r3HJglucosamine and (?H]fucose was used (LAVOIEet a/., 1978), permits the interpretation that proteins can migrate as far as the Go&i apparatus in the presence of cobalt ions. it follows that cobalt is acting after fu&osy~tio~~ at some step during the transfer of fast-transported proteins to the transport system. Analogies hetweet! neurons and other secretory cells

The presence of a cobalt-sensitive, calcium-dependent step during this transfer suggests an analogy between processing of fast-transported proteins by neuronal somata. and processing of secretory proteins by exocrine cells. Both fast transport and secretion involve the sorting out of a subclass of proteins for export from the soma Pulse-chase studies of nerve cells. monitoring the positions of radj~uto~ap~~ grains at successive intervals after protein synthesis (DROZ, 1965: Drtoz & KOENIG,1970), have suggested that proteins being processed for fast axonal transport follow a route through rough endoplasmic reticulum and Golgi apparatus similar to that of the wellstudied secretory proteins of pancreatic exocrine cells (PALADE, 1975; JAMESON,1975). The-passage of secretory proteins to the plasma membrane occurs within zymogen granules which form from precursor

1199

organelles that bud off from the Golgi apparatus. A formal analogV between the initiation of fast axonal transport and secretion must then include a proposal that proteins destined for fast transport also become packaged in ~br~e-~mited granules derived from the Golgi apparatus. The calcium requirement observed during initiation of transport (HA-LAG et at, 19%) could then involve delivery of the proteins to the transport system in a manner similar to that whereby calcium is generally required to tripger the release of secretory products (Dou~u$ 1968; RUBIN, 1970) Thus, the decrease in the amount of protein that undergoes fast axonal transport during incubation of ganglia in ~l~urn-free medium (HAMMERSCH~AG et a!., 19753, and the inhibition of enzyme secretion from exocrine pancreas that results from the in vitro removal of extracellular calcium (HOKIN. 1966; HEISLER,FAST & TENENE#~W~,1972; KA~~JNO. 19721 may have a similar etiology. It is of interest that the major source of calcium for pancreatic enzyme release is proposed to be intra~ilular stores rather than the extracellular medium (CASE & CLAUSEN, 1973; WILLIAMS& CI~ANDUR,1975). Since the addition of Co’+ to a calcium-free medium depresses the initiation of transport to a greater extent than does calcium-free medium alone (HAMMERXXLAG et at, 19761, it appears that the calcium required for this process is also supplied from an intracellular pool. Such a mechanism differs from that described for secretion from the endocrine pancreas+ adrenal medulla, neur~~p~ysis and nerve terminals (DOUGLAS 1968; RU~IN, 19701, which clearly depends on the influx of ex~a~llular calcium for release of the particular hormones or neurotransmitters. In addition, for each of these endocrine-type release mechanisms, Sr”+ or Ba2 + act as Ca2+ agonists while Co2+ or Mg2+ act as antagonists. By contrast, Sr* + but not I%‘+, and Co2+ but not release ~~~~A~~ & Mg=+, affect exocrine CHANDLE&1975) and initiation of fast axonal transport (HAMt#X!%Xi LAG et cd., 19”/7). While this analogy predicts a Golgi apparatusassociated packaging of fast-transport proteins, it is clear that neoronal somata do not contain membrane-bound granules similar to the relatively large zymogen granules characteristic of pancreatic a&tar cells (JAMESON,1975). Neurons, however, do contain large numbers of heterogeneous smaller granules associated with their more diffuse Golgi apparatus (Prrmas, PAUY & WEEXER, 1976). It is also of interest that after pancreatic tissue is stimulated continuously for several hours, secretory proteins become packaged, prior to release, in storage granules considerably smaller than the usual zymogen granules (JAMIESON & PAWDE, 1971). These morphological dirferences are likely to be related to fun~onal demand. Digestive enzymes are normally secreted in high-level int~i~ent bursts, whereas fast-~sport proteins appear to be exported in a more continuous manner. Thus, intracellular events comprising initiation of fast

R. HAMMERSCHLAG and P.-A. LAVOIE

1200

axonal transport may resemble more closely the processing of secretory proteins in the continuously stimulated pancreatic cell. If fast transport of protein occurs in association with the axonal network of smooth endoplasmic reticulum, as some current evidence suggests (HOLTZMAN, 1971; BYERS, 1974; SOTELO& RICHE, 1974; DROZ, RAMBOURG & KOENIG, 1975 ; MARKOV, RAMBOURG& DROZ, 1976), entry of proteins into the smooth reticular lumen would be topologically equivalent to the transfer of secretory proteins out of the cell (TRUMP, 1975: ROTHMAN& LENARD,1977). Alternatively, fast transport may occur in association witi the membrane of the reticulum and not within the lumen (see

DROZ, 1975). In this case, the final step in the transfer of proteins to the transport system might then be similar to the intracellular process by which proteins are inserted into the plasma membrane (ROTHMAN& LENARD,1977) rather than secreted through it.

Acknowledgements-The present investigation was supported by NSF Grant BNS75-17640 (R.H.), by a MRC of Canada Fellowship to P.-A.L.. and by the Kenneth Cory Research Fund administered through the City of Hope National Medical Center. We thank Dr JAMESE. VAUGHN for his critical reading of the manuscript. and FRANCI B~LEN. ALBERT TJAN and ROBERT OXPER for their valuable assistance in the completion of these experiments.

REFERENCES R. T. & TREISTMAN S. N. (1977) Glycoproteins are modified in the axon of R2. the giant neuron of Apl.~,sitr cahfornica, after intra-axonal injection of [‘H]N-acetylgalactosamine. Brain Res. 121, 287-309. BARKER J. L., HOFFMAN P. N.. GAINER H. & LASEK R. J. (1975) Rapid transport of proteins in the sonic motor system of the toadfish. Brain Res. 97, 291-301. BARONDES S. H. (1970) Brain glycomacromolecules and interneuronal recognition. In The Neurosciences: Second Stud! Program (ed. SCHMIDTF. 0.). pp. 747-760. Rockefeller University Press, New York. BENNETTG., GIAMBERARDINO Dr L., KOENIG H. L. & DROZ B. (1973) Axonal migration of protein and glycoprotcin to nerve endings--II. Radioautographic analysis of the renewal of glycoproteins in nerve endings of chicken ciliary ganglion after intracerebral injection of [3H]fucose and [3H]glucosamine. Bruin Res. 60, 129-146. BYERSM. R. (1974) Structural correlates of rapid axonal transport: evidence that microtubules may not be directly involved. Brain Res. 75, 97-113. CASE R. M. & CLAUSENT. (1973) The relationship between calcium exchange and enzyme secretion in the isolated rat pancreas. J. Physiol., Land. 235, 75-102. DOUGLASW. W. (1968) Stimulus-secretion coupling: the concept and clues from chromaffin and other cells. Br. J.

AMBRON

Pharmac.

34, 451-474.

DRAVID A. R. & HAMMERSCHLAG R. (1975a) Axoplasmic transport of proteins in oitro in primary afferent neurons of frog spinal cord: effect of Ca “-free incubation conditions. J. Neurochem. 24, 71 l-718. DRAVIDA. R. & HAMMERSCHLAG R. (1975b) The role of calcium in axonal transport: gel electrophoretic comparison of proteins undergoing ‘fast’ transport in normal and calcium-free medium. Abstr. Int. Sot. Neurochem. 5, 261. DROZ B. (1965) Accumulation de protiines nouvellement synthktides dans l’appareil de Golgi du neurone: ttude radioautographique en microscopic tlectronique. CR. hebd. S&K. Acad. Sci., Paris 260, 320-322. DROZ B. (1975) Synthetic machinery and axoplasmic transport: maintenance of neuronal connectivity. In The Nervous System. Vol. 1: The Basic Neurosciences (ed. BRADYR. 0.). pp. I1 I-127. Raven Press, New York. DROZ B. & KOENIGH. L. (1970) Localization of protein metabolism in neurons. In Prorein Metabolism of the Nervous Sysrem (ed. LAJTHAA.), pp. 93-108. Plenum Press, New York. DROZ B., RAMBOURGA. & KOENIGH. L. (1975) The smooth endoplasmic reticulum: structure and role in the renewal of axonal membrane and synaptic vesicles by fast axonal transport. Brain Res. 93, l-13. EDSTR~MA. & MATT~.WNH. (1972) Rapid axonal transport irt oilro in the sciatic system of the frog of fucose-. glucosamine- and sulfate-containing material J. Neurochem. 19, 1717-1729. EDSTR&WA. & MACON H. (1973) Electrophoretic characterization of leucine-, glucosamine- and fucose-labelled proteins rapidly transported in frog sciatic nerve. J. Neurochem. 21. 1499-1507. FENG T. P. & LIU Y. M. (1949) The connective tissue sheath of the nerve as effective diffusion barrier. J. cell. camp. Physiol.

34. l-16.

FORMAND. S.. GRAFSTEINB. & MCEWEN B. S. (1972) Rapid axonal transport of [‘Hlfucosyl glycoproteins in the goldfish optic system. Brain Res. 48, 327-342. FRANKENHAEUSER B. (1957) The effect of calcium on the myelinated nerve fibre. J. Physiol., Land. 137, 245-260. GRAFSTEIN B. (1977) Axonal transport: the intracellular traffic of the neuron. In Handbook of Physiology Sect. 1. Vol. I: Cellular Biology of Neuroses (ed. KANDEL E. R.), pp. 691-717. Amer. Physiol. Sot., Bethesda, Maryland. HAMMERSCHLAG R., BAKHIT C., CHIU A. Y. & DRAVIDA. R. (1977) Role of calcium in the initiation of fast axonal transport of protein: effects of divalent cations. J. Neurobiol. 8, 43945 1. HAMMERSCHLAG R., CHIU A. Y. & DRAVIDA. R. (1976) Inhibition of fast axonal transport of [‘HIprotein by cobalt ions. Brain Res. 114, 353-358. HAMMERSCHLAG R., DRAVIDA. R. & CHIU A. Y. (1975) Mechanism of axonal transport: a proposed role for calcium ions. Science, N.Y. 188, 273-275. HEISLERS., FAST D. & TENENHOU~EA. (1972) Role of Ca*+ and cyclic AMP in protein secretion from rat exocrine pancreas. Biochim. biophys. Acta 279, 561-572.

Initiation of fast axonal transport

1201

HESLOPJ. P. (1975) Axonal flow and fast transport in nerves. Adu. camp. Physic/. Biochem. 6, 75-163. HOKINL. E. (1966) Effects of calcium omission on acetylcholine-stimulated amylase secretion and phosphohpid synthesis in pigeon pancreas slices. &o&m. biophys. Acta 115, 219-221. HOLTZMANE. (1971) Cytochemical studies of protein transport in the nervous system. Phil. Trans. R. Sec. Ser. B 261, 407-421. JAMIESONJ. D. (1975) Membranes and secretion. In Cell Membranes: Biochemistry, Cell Biology and Pathology (ads WEISSMANN G. & CLAI~~RNER.), pp. 143-152. H.P. Publishing Co., New York. JAMIES~NJ. D. & PALADEG. E. (1971) Synthesis, intracellular transport. and discharge of secretory proteins in stimulated pancreatic exocrine cells. J. Cell Biol. 50, 135-158. KANNO T. (1972) Calcium-dependent amylase release and electrophysiological measurements in cells of the pancreas. J. Physioi., Land. 226, 353-371. KARLSSON J.-O. & SJ&TRANDJ. (1971) Rapid intracellular transport of fucose-containing glycoproteins in retinal ganglion cells. J. Neurochem. 18, 2209-2216. KRNJEVICK. (1954) Some observations on perfused frog sciatic nerves. J. Physiol., Lond. 123, 338-356. LANGLEY0. K. & CARMJCHAEL N. G. (1978) Glycoprotein transport in the rat cerebellum. J. Neurocytol. 7, 479488. LAVOIEP.-A.. BOLEN F. & HAMMERSCZKAGR. (1979) Divalent cation specificity of the calcium requirement for fast transport of proteins in axons of desheathed nerves. J. Neurochem. In press. LAVOIEP.-A., HAMMERSCHLAG R. & TJAN A. (1978) Cobalt ions inhibit fast axonal transport of [3H]glycoproteins but not glycosylation. Brain Res. 149, 535-540. LEBLONDC. P. & BENNETT G. (1977) Role of the Golgi apparatus in terminal glycosylation. In Inrernariomd Cell Biology: 19761977 (eds BRINKLEY B. R. & PORTERK. R.), pp. 326-336. Rockefeller University Press. New York. MARKO P. & CU~NODM. (1973) Contribution of the nerve cell body to renewal of axonal and synaptic glycoproteins in the pigeon visual system. Brain Res. 62, 419-423. MARKOVD., RAMBOURG A. & DROZ B. (1976) Smooth endoplasmic reticulum and fast axonal transport of glycoproteins. an electron microscope radioautographic study of thick sections after heavy metals impregnation. J. Microsc. Biol. Cell. 25, 57-60. OCHS S. (1974) Axoplasmic transport-Energy metabolism and mechanism. In The Peripheral Neruous System (ed. HUBBARDJ. I.), pp. 47-72. Plenum Press, New York. OCHS S.. WORTH R. M. & CHAN S.-Y. (1977) Calcium requirement for axoplasmic transport in mammalian nerve. Nature, Land. 270, 748-750.

PALADEG. (1975) Intracellular aspects of the process of protein synthesis. Science, N.Y. 189, 347-358. PETERSA., PALAYS. L. & WER~TERH. de F. (1976) The neuronal cell body. In The Fine Structure oj the Nervous System: The Neurons and Supporring Cells, pp. 11-65. W. B. Saunders, Philadelphia ROTHMANJ. E. & LENARDJ. (1977) Membrane asymmetry. Science, N.Y. 195, 743-753. RUBINR. P. (1970) The role of calcium in the release of neurotransmitter substances and hormones. Pharmac. Reu. 22, 389-428. SOIELO C. & RICHE D. (1974) The smooth

endoplasmic reticulum and the retrograde and fast orthograde transport of horseradish peroxidase in the nigrostriatal-nigral loop. Anat. Embryo/. 146, 209-218. SPIRO R. G. (1970) Glycoproteins. A. Rev. Biochem. 39, 599-638. TRUMPB. F. (1975) The network of intracellular membranes. In Cell Membranes: Biochemistry, Cell Biology and Pathology (eds WEI~~MANG. & CLAIBORNER.), pp. 123-133. H.P. Publishing Co., New York. WILLIAMSJ. A. & CHANDLER D. (1975) Ca2+ and pancreatic amylase release. Am. J. Physiol. 22& 1729-1732. ZATZ M. & BARONDESS. H. (1971) Rapid transport of fucosyl glycoproteins to nerve endings in mouse brain. J. Nemo&em. 18, 1125-l 133. (Accepted 19 February

1979)