Role of gangliosides in the uptake and retrograde axonal transport of cholera and tetanus toxin as compared to nerve growth factor and wheat germ agglutinin

Brain Research, 132 (1977) 273-285 t) Elsevier/North-Holland Biomedical Press

273

ROLE OF G A N G L I O S I D E S IN T H E U P T A K E A N D R E T R O G R A D E A X O N A L T R A N S P O R T OF C H O L E R A A N D T E T A N U S T O X I N AS C O M P A R E D TO NERVE G R O W T H F A C T O R A N D W H E A T G E R M A G G L U T I N I N

K. STOECKEL, M. SCHWAB and H. THOENEN Department o/Pharmacology, Biocenter O/the University, CH-4056 Basel (Switzerland)

(Accepted December 17th, 1976)

SUMMARY Previous investigations have shown that tetanus toxin is transported retrogradely in all peripheral neurons whereas the transport of N G F is confined to adrenergic and sensory neurons. Other macromolecules with molecular weights and general physiochemical properties similar to N G F and tetanus toxin (e.g., cytochrome C, insulin, horseradish peroxidase and bovine serum albumin) are not transported to a detectable extent if injected in comparable molar concentrations. For tetanus toxin, which is transported in all peripheral neurons, it has to be assumed that it's retrograde transport depends on properties common to all neurons. In view of the relatively high ganglioside content of the neurons and the high affinity of tetanus toxin for the trisialoganglioside GT1, we studied the influence of gangliosides on the retrograde transport of tetanus toxin as compared to NGF. We included into the study cholera toxin which is known to have a high affinity for the monosialoganglioside G~I and wheat germ agglutinin, a lectin with specific affinity for glycoproteins with N-acetylglucosamine residues. Both cholera toxin and wheat germ agglutinin were transported efficiently in all peripheral neurons. Preincubation of 12'~l-cholera toxin with monosialoganglioside GM1 completely blocked its retrograde axonal transport. The transport of N G F and wheat germ agglutinin was affected neither by various purified gangliosides nor by a mixture of bovine brain gangliosides. The transport of tetanus toxin was only reduced by 50%, both by the trisialoganglioside GT1 and the bovine ganglioside mixture.

I NTROD UCTION In previous studies it has been shown that tetanus toxin is taken up by the nerve terminals of all peripheral neurons (motor, sensory, and adrenergic) and is transported

274 retrogradely to the corresponding cell bodies ~'~. In contrast, the retrograde axonal transport of nerve growth factor ( N G F ) is confined to sensory and adrenergic neurons 2~ and that of antibodies to dopamine fl-hydroxylase exclusively to adrenergic neuronslL The rate of retrograde transport in a given species of neurons is the same for all macromolecules transported. Moreover, N G F and tetanus toxin are localized in the same cellular compartments of axons and cell bodies as far as can be judged from electron microscopic autoradiography 2t and histochemical localization of coupling products between N G F and horseradish peroxidase ~'~. These observations suggest that the uptake of a macromolecule depends on its affinity to binding sites on the membrane of the nerve terminal and that this binding initiates the uptake into the nerve terminal and the subsequent transport with a carrier which is identical for all macromolecules transported. After demonstrating that not only tetanus toxin but also cholera toxin i.~ transported retrogradely by all peripheral neurons, we investigated whether the common binding sites for these two toxins in all nerve terminals are gangliosides, since cholera toxin has been reported to have a high affinity for the monosiatoganglioside GMI a,~,1~-~6 and tetanus toxin for the trisialoganglioside Ga,ll,L We wish to report that the monosialoganglioside Grit completely blocked the retrograde axonal transport of cholera toxin, whereas the retrograde transport of tetanus toxin was reduced only by 50 °/o by both purified trisialoganglioside G,~-~and a mixture of bovine brain gangliosides. In contrast, the retrograde ax onal transport of N G F was not impaired at all and nor was that of wheat germ agglutinin, a macromolecule transported retrogradely in all peripheral neurons. MATERIALS AND METHODS N G F was isolated as the 2.5 S subunit according to the procedure of Bocchini and Angeletti 3. The purity of N G F was controlled by SDS gel electrophoresis and its biological activity was determined according to Fenton 1° in a tissue culture assay with dorsal root ganglia of 7-9-day-old chicken embryos. The biological activity ranged between 220 and 260 biological units (BU) per/~g of protein. Tetanus toxin was a gift from Drs. Bizzini and Turpin. lnstitut Pasteur. Paris. The specification details of their preparation has been reported recentlyL Cholera toxin was purchased from Schwarz and Man. New York. Wheat germ agglutinin was a gift from Dr. M. M. Burger. Biozentrum. Basel. A detailed characterization of this lectin which binds specifically to N-acetylglucosamine residues has been reported recently".

Gangliosides Ganglioside mixture. The ganglioside mixture used was an extract from bovine brain and purchased from Koch-Light Laboratories. Bucks. Great Britain. Before use the ganglioside extract was dissolved in distilled water at a concentration of 2 mglml and centrifuged for 20 rain at 9000 × g. The pellet formed amounted to about 5')/o of the total ganglioside mixture. One ml samples of the supernatant were frozen,

275 lyophilized and stored at --40 °C. The composition of the ganglioside mixture analyzed by thin layer chromatography (TLC) (details see below) was about 50 % GMI, 20~,/, GM2, 10}~ GraB. In addition the TLC showed two unidentified non-polar components which represent about 5 °/o each. For injection 1.9 mg of the ganglioside mixture were incubated for 30 min with 150 #1 of 0.05 M phosphate buffer containing about 40 #g of l"q-labeled protein (tZSl-wheat germ agglutinin, 125I-NGF or x2'51cholera toxin). Before use the incubation mixture was centrifuged at 100,000 >< for 30 rain but no further precipitation could be detected. PurOqed gangliosides GM1, GTI. The individual gangliosides G_ul and Gaq were purchased as organic solvent solutions (chloroform-methanol) from Supelco, Belefonte, Pa. (U.S.A.). Before use the ganglioside preparations were controlled by TLC (Merck, TLC silica gel plates 60 Fz54). As solvent system n-propanol-water (7:3 by volume) was used. The spots were visualized either by 50 3o sulfuric acid spray and heating or by spraying with resorcinol ~7. The quantitative composition of the ganglioside preparations was analyzed by light scanning of photographs of the thin layer chromatograms. The purified GM1 contained more than 95 ~£iGM1 together with a small amount of GM2. The purified GT1 preparation was a mixture containing about 50 ?,/, GT1, 40 ~, GD1B, 5 ~(/~GM1 and 5 ?i, G.w,. Before injection 100/~1 of a solution of 500 #g of purified gangliosides in organic solvent were evaporated under a mild stream of nitrogen and then dissolved in 150/~1 of phosphate buffer 0.05 M, pH 7.8, containing about 40 /zg of the lZSl-labeled proteins. Incubation and centrifugation of the samples were performed as described above. The injection procedure is described below.

Labeling of the proteins The labeling of the different proteins (NGF, cholera toxin, tetanus toxin, wheat germ agglutinin) was performed with Naleq and chloramine T as described in detail previously '~3. Nal')'~l was purchased from EIR Wiirenlingen, Switzerland at a specific activity of 8-15 Ci/mg iodine. For each preparation 200/~g of protein and about 5 mCi of Na~'~6I were used. The specific activity of the labeled proteins were 15 ~tCi//~g for le'q-cholera toxin, 8 #Ci//~g for v"51-NGF, 20 ~Ci/#g for 12q-wheat germ agglutinin and 7/,Ci/¢tg for 12'~l-tetanus toxin.

Anima]s For all experiments female Sprague-Dawley rats, weighing 200-250 g, were used. The rats were kept at a constant temperature (23 °C), supplied with the usual lab chow diet (NAFAG AG, Gossau, SG, Switzerland) and tap water ad libitum.

Injection procedure The injection procedure used for studying the retrograde axonal transport in the peripheral nervous system has been described in detail before 19. Briefly, for studying the retrograde transport in adrenergic neurons the labeled macromolecules were injected unilaterally into the anterior eye chamber. The superior cervical ganglia of the injected and non-injected side were removed 2-48 h later. The

276 difference in the accumulation of radioactivity between the ganglia was taken as a measure of retrograde transport. The same principle was used for studying the retrograde transport in sensory and motor neurons. For the sensory neurons the site of injection was the forepaw and the time-course of the difference of accumulation of radioactivity was determined in the spinal ganglia C6 and C7. For the motor neurons the injection site was the deltoid muscle and the side difference in the accumulation of radioactivity in the spinal cord was determined in segment C6-Cs. The injection volume for the anterior eye chamber was always l0/~1, for the forepaw, 20 #1, and the musculus deltoideus. 30 ul. The radioactivity accumulated in the ganglia and the spinal cord segments was determined by counting the samples directly in a Packard y-counter model 3001 at a counting efficiency of 65" . All values are given in counts/min and are not corrected for the efficiency.

A utoradiographic studies For the autoradiographic studies t0 ~1 of ~251-cholera toxin or ~eSl-wheat germ agglutinin were injected unilaterally into the anterior eye chamber and 30 #1 into the submandibular gland of adult rats. Fourteen hours after injection the animals were fixed by perfusion through the heart with 2.5'~ glutaratdehyde and t ','(, formaldehyde in 0.1 M phosphate buffer, pH 7.4. postfixed in 1.33'7o OsO,t, dehydrated and embedded in Epon 812. Semithin sections were mounted on glass objective slides. coated with llford L4 emulsion diluted 1:1 with distilled water, and e x p o ~ d for 30 days at 4 °C. Autoradiograms were developed with Kodak D I 9 developer at 20 C for 6 rain and fixed in 24 o/sodium thiosulfate. For electron macroscopic autoradiography 70 nm sections were mounted on collodion coated microscope slides, dipped into llford L4 emulsion diluted I :3 with distilled water, dried and exposed in the dark for 3 months at 4 °C. Autoradiograms were developed with Microdol-X for 3 rain at 20 ~C. fixed with 24 },J, sodium thiosulfate (2 min and stained with uranyl acetate and lead citrate. RESULTS

Retrograde axonal transport of cholera toxin Adrenergic neurons. After unilateral injection of VZSl-labeled cholera toxin into the anterior eye chamber there was a highly preferential accumulation of radioactivity in the superior cervical ganglia of the injected side (Fig. 1). However. this difference did not become apparent before 6 h. Up to 4 h after injection there was no statistically significant (P :> 0.05) difference between the injected and non-injected side. F r o m the time-lag between the injection of 125I-cholera toxin and the appearance of a statistically significant difference m the radioactivity between the injected and non-injected side a rate of retrograde axonal transport of about 3 mm/h can be calculated for a distance of about t 5-20 m m between the adrenergic nerve terminalsin the iris and the corresponding cell bodies in the superior cervical ganglia. This rate of retrograde axonal transport is very close to that calculated for N G F 13. tetanus toxin 26 and antibodies to dopamine fl-hydroxylase 11.

277 injecfed

~ SO0

i---I non-

side

mjecfed

t

side

g 200 c E

100 0

! "-,~! . -II . . . . . . . . . 1

I

2

&

r

6

! I

• .........

~

I

8

I0 12 14 16 1

hrne

I

.....

I

affer

i

injecfion

Io

i_.,,//---!

,

2

I

If'

22 Z/*

/.8

,n hours

Fig. 1. Time-course of accumulation of radioactivity in the superior cervical ganglion after unilateral injection of 35 #Ci of l~al-labeled cholera toxin into the anterior eye chamber of 200 250 g female rats. The animals were killed 2-48 h after injection. The superior cervical ganglia of the injected and noninjected side were dissected and the radioactivity directly determined in a y-counter. Each point represents the mean ~ S.E.M. (n ~ 5 7).

Sensor), neurons. Fig. 2 shows the time-course of accumulation of radioactivity in dorsal root ganglia C6 and C7 after unilateral injection of 12q-cholera toxin into the forepaw. The first statistically significant (P < 0.05) difference between the injected and non-injected side occurred after 4 h. Using the same criteria as in the adrenergic neurons an approximate rate of retrograde transport of 13 mm/h can be calculated. This rate of transport is very similar to that determined for N G F and tetanus toxin in previous experiments25, 96. Motoneurons. After unilateral injection of lz~I-cholera toxin into the musculus deltoideus a statistically significant (P < 0.05) difference in the accumulation of radioactivity in the spinal cord was present after 6 h (Fig. 3). Assuming an average distance of about 35 mm between the site of injection and a lag-phase of 5 h for the arrival of the first moiety of retrogradely transported r~5l-cholera toxin a rate of retrograde transport of about 7 mm/h can be calculated. This rate of transport corresponds to that calculated for tetanus toxin in previous experiments% T

injected s~de

160O h--'l non-injected

T

I . ~

side

~~~ w+oo 1200

%

~

/I /

1000 BOO

/i

600 c 400

tf/!

i

2

,

4

i

6

,

8

i

i

i

r

r

J

i

i

lO 12 II 16 18 20 22 21+ t,me

after

injecfion

i/

r

48

in hours

Fig. 2. Time-course of accumulation of radioactivity in the dorsal root ganglia C6, C7 after unilateral injection of 70 #Ci of 12q-labeled cholera toxin into the forepaw of 200-250 g rats. At the given time points the animals were killed, the dorsal root ganglia of the injected and non-injected side were removed and their radioactivity determined in a y-counter. Each point represents the mean h S.E.M. (n

5 7).

278

~

7000

L~

6000

njecreo

Side

5000 ~000

300C

2000 E

i t -l ~ i---l...,

•

000 i

l

?

I

I

6

~

l

I

I

I

I

10 12 lt. 16 18 20 22 t me after nje~rlon - h o u r s

I

2/,

4/-

/~. " ' ~ 48

Fig. 3. Time-course of accumulation of radioactivity in the spinal cord segment C0-C8 after unilateral injection of 90/~Ci of ~5l-labeled cholera toxin into the musculus deltoideus of 200-25Og female rats. The rats were killed 2-48 h after injection. The spinal cord segments C6-C8 were removed and divided mediosagittally. The radioactivity in the two halls were determined directly in a y-counter. Each point represents the mean ~. S.E.M, (n 6-7).

Retrograde axonal transport of wheat germ agglut&in After preliminary experiments with concanavatin A. p h y t o h e m a g g l u t i n i n and wheat germ agglutinin had s h o w n that wheat germ agglutinin was the m o s t efficiently transported lectin, we studied its retrograde transport in all three neurons. Fig. 4 shows that after 16 h there was a marked difference between the injected and noninjected sides in all neurons studied.

Effect of gangtiosides on the retrograde axonal transport of maeromolecules Colera toxin. It has been s h o w n under various experimental conditions that the activation o f adenylate cyctase by cholera toxin depends on the binding o f its Bsubunit to the m o n o s i a l o g a n g l i o s i d e Gra~ o f the corresponding cell membrane 5-te,16 In order to obtain information as to whether this ganglioside might also act as a injecle~ SlOe

l

"--Inon

_

nlected

SlOe

8000-//.4 Z/A //A

600( F

~/A //.4

aOOC--

/A/AZ/A

E

c~

~fA

//A //A

))~

f//

//A

2000--

H.//'4 //A //A

adrenerglc

sensory

mo';oneuron$

Fig. 4. Accumulation of radioactivity m superior cervical ganglia (adrenergic), dorsal root ganglia C0, C7 (sensory) and spinal cord segment C~-Cs 16 h after unilateral injection of 12St-labeled wheat germ agglutinin. Forty I~Ciof 12SI-wheat germ agglutinin were injected into the right anterior eye chamber (adrenergic), 70 uCi into the right forepaw (sensory) and 90/~Ci into the muscutus deltoideus (motoneurons). The accumulated radioactivity on the injected and non-injected side was directly determined in a ~,-counter. Each column represents the mean -+. S.E.M. (n 5 7).

279

injected side

_g 700t &

-

//

I

] non- injected side

//

500L c~

9300c

E

1oo -

ChT confrol

ChT IJTI{GMI}

ChT GM~

Fig. 5. Effect of gangliosides on the retrograde axonal transport of 12q-cholera toxin in adrenergic neurons. The accumulation of radioactivity was measured in the superior cervical ganglia 14 h after unilateral injection of 35/tCi lZq-cholera toxin (control) and 35/tCi ~2~I-choleratoxin preincubated with either 35 #g of partially purified trisialoganglioside G,m or 35 #g of purified monosialoganglioside GM~ into the anterior eye chamber. The radioactivity in the superior cervical ganglia was directly determined in a ),-counter. Each column represents the mean ± S.E.M. (n 5-7). receptor, and thus provide the prerequisite for uptake and retrograde axonal transport, l"5I-labeled cholera toxin was incubated for 30 rain with either GM~ or partially purified GT1 ganglioside. In order to eliminate the possibility that the interference of these gangliosides with the retrograde axonal transport might result from the formation of precipitates the incubation mixture was subjected to a 30 rain centrifugation at 100,000 × g immediately before injection. Preincubation with G ~ completely blocked the retrograde axonal transport of t25I-labeled cholera toxin (Fig. 5). Preincubation with the 'purified' (see Methods) trisialoganglioside GTI produced a reduction of about 50 ~!~;.However, the TLC analysis of'purified' GT1, revealed that it contained 40 °Jo GDII3, and each 5 % G~[1 and GM2. Therefore, the blocking effect could well be due to GM1 contamination. Attemps to further purify GT1 by preparative TLC were not successful on account of the low stability of GT1. Rechromatography of the purchased 'purified" GT1 again revealed a mixture of different hydrolysis products including G~a. Thus, even the incubation of cholera toxin with freshly prepared Gaq would not exclude the formation of GM1 in the tissue which then could be responsible f o r the partial blockade. Tetanus toxin. Van Heyningen reported that tetanus toxin binds both to the diand trisialogangliosides GraB and GT1 ~5. As shown in Fig. 6 neither preincubation of leq-tetanus toxin with a mixture of bovine brain gangliosides nor with partially purified GT~ completely abolished the retrograde axonal transport of this toxin. Both preincubation with GTt and the ganglioside mixture resulted at the best in a reduction by 40~(,. Moreover, preincubation with G m did not at all impair the retrograde transport of tetanus toxin.

280 r/~/~/~ iDjecteo sloe

2400 --

ooo[ g

r-'7

non - mjecteo s~cle

~z

~/A ~ 1600L.--

,5

~A

//

7200 rI

"/A ,//'/

~A ;/.,6;

k~ ~oo~//1

E ~oo~TT

COr~troI

gangtlos~des

TT ~T~

h_

TT

'] M.

Fig. 6. Effect of gangliosides on the retrograde axonal transport of ~zS-tetanus toxin in adrenergic neurons. Twenty l~Ci of le'~l-tetanus toxin fcontrol) and 20/~Ci ~2~[-tetanu8 toxin together with 125 Hg bovine brain ganglioside extract, 35/tg purified trisialoganglioside GT1 and 35/~g purified monosialogangiioside G.~I1,respectively, were injected unilaterally into the anterior eye chamber of 200-250 g female rats. The superior cervical ganglia of the injected and non-injected side were removed 14 h after injection. The radioactivity accumulated in the ganglia was directly measured in the y-counter. Each column represents the mean :r S.E.M. (n ~ 5-6). njectec] sloe Z400 -

r'7

c o

non-injefteci sloe

zooo -

16oo 3200

VA

YA 7A

-

/,

BOO -

400 -

NGF

control

NGF

gangllOSldeS

NGF

G~

NGF GM

Fig. 7. Effect of gangliosides on the retrograde axonal transport of I~51-NGF in adrenergic neurons. Twenty #Ci 12q-NGF (control) and 20 uCi 1251-NGF together with 125 ~tg bovine brain ganglioside mixture, 35 ug purified trisialoganglioside GT1 and 35 /*g purified monosiaioganglioside GMt, respectively, were injected unilaterally into the anterior eye chamber of 200-250 g female rats. The accumulated radioactivity in the superior cervical ganglia was measured 14 h after the injection by removal of the ganglia and their direct counting in a y-counter. Each column represents the mean :xS.E.M. (n - 6-7). The high proportion of radioactivity on the non-injected side coming from t~5I-NGF, which reaches the nerve terminals of the contralateral side via the general circulation, has been extensively discussed in a previous paper 2a.

N G F and wheat germ agglutinin. N o d a t a a r e a v a i l a b l e i n d i c a t i n g t h a t N G F o r w h e a t g e r m a g g l u t i n i n b i n d to g a n g l i o s i d e s . A c c o r d i n g l y , t h e i r r e t r o g r a d e a x o n a t t r a n s p o r t was n o t i m p a i r e d by a m i x t u r e o r by p u r i f i e d g a n g l i o s i d e s (Figs. 7, 8).

Autoradiography Light microscopic autoradiograms

o f rat s u p e r i o r c e r v i c a l g a n g l i a 14 h a f t e r

281 injected side o

~ /

sooo7

""

-

I---I non

inJected Side

) 4000 -

2

y/,,

2ooo-

#

WGA control

WGA gangllosides

Fig. 8. Effect of a mixture of bovine brain garlgliosides on the retrograde axonal transport of Is~Ilabeled wheat germ agglutinin in adrenergic neurons. The accumulation of radioactivity was measured in the superior cervical ganglia 16 h after the unilateral injection of either 50 pCi 1~SI-wheat germ agglutinin or 50/~Ci ~251-wheat germ agglutinin together with ]25 l~g bovine brain ganglioside mixture. The ~-°'~]-wheat germ agg]utinin was injected transscleral]y into the anterior eye chamber of 200-250 g rats. The removed ganglia were directly counted in a 7-counter. Each column represents the mean ± S.E.M. (n = 6-7).

unilateral injection of 1sSI-cholera toxin into the anterior eye chamber and the submandibular gland revealed medium to heavy labeling of a limited number of adrenergic neurons (Fig. 9a). Correspondingly, single labeled fibers could be observed in the postganglionic nerve. The glia and the majority of the neurons were unlabeled. A similar selective labeling of a small population of adrenergic neurons, presumably corresponding to cells innervating the iris and the submandibular gland, was observed 14 h after injection of 1251-wheat germ agglutinin (Fig. 9b). In this case the labeling intensity over individual neurons was very high. Electron microscopic autoradiograms of postganglionic sympathetic axons revealed an exclusive localization of 12aI-cholera toxin and leSl-wheat germ agglutinin inside the axons (Fig. 10), as shown previously for lesI-NGF and 1251-tetanus toxin2~,22. DISCUSSION After previous studies had shown that the retrograde axonal transport of N G F is restricted to adrenergic and sensory neurons 26 whereas tetanus toxin is transported in all peripheral 26 and central neurons 20 investigated so far, the present study revealed a retrograde transport of two additional macromolecules, namely cholera toxin and wheat germ agglutinin, which are also transported in all peripheral neurons. Light and electron microscopic autoradiograms provided direct evidence for the intraaxonal transport of these two molecules. The retrograde transport of cholera toxin, wheat germ agglutinin and tetanus toxin by all peripheral neurons suggests that the uptake and retrograde axonal transport of these macromolecules depends on properties common to many or all

282

Fig. 9. a : autoradiogram (dark-field illumination) of a semithin section through a rat Su~c~:ior ~:crvica} ganglion 14 h after injection of ~e~l-cholera toxin into the anterior eye chamber (10 ,i,I1 and the sub° mandibular gland (3 • I0/d). b: autoradiogram (phase contrast) of a section through a ral superior cervical ganglion 14 h after injection of 12n|-wheat germ agglutinin into the anterio]: eye chambe~ (10/d) and the submandibular gland (3 10/,I). Magnification 1()0.

283

Fig. 10. Electron microscopic autoradiograms of postganglionic axons in the rat superior cervical ganglion 14 h after injection of 12'~l-choleratoxin (a) or r-':'l-wheatgerm agglutinin (b) into the anterior eye chamber and the submandibular grand. Magnification x 11,000. neurons. It is known that gangliosides represent a relatively large proportion of neuronal membrane lipids7, s,17 and that both cholera toxin and tetanus toxin are bound with high affinity to gangliosides 4,5,11-1~. We therefore investigated whether the retrograde axonal transport of these two proteins is linked to the presence of gang[iosides in the neuronal membrane. Indeed, the retrograde axonal transport of cholera toxin was completely abolished by a crude mixture of bovine brain gangliosides and a virtually pure fraction of GM1 ganglioside, to which cholera toxin has been shown to bind with high affinity 4,'~,14-1~. The partial blockade of the retrograde transport of cholera toxin by the purified trisialoganglioside GT1 could be explained by the fact that this ganglioside is not very stable, and a continuous hydrolysis to G•I has been shown to occur during chromatographic separation. It has also been shown that the high affinity binding of cholera toxin to GM1 in confined to the 66,000 molecular weight binding unit's, 14. However, according to Cuatrecasas et al. 6 the labeling of the native cholera toxin with ~2~I occurs almost exclusively in the catalytic unit (molecular weight 36,000). An investigation of the accumulated radioactive compound in the superior cervical ganglion by SDS gel electrophoresis was therefore unsuitable to answer the important question as to whether, after binding to the cell membrane, only the catalytic unit or both, the binding and catalytic unit are transported retrogradely. It is noteworthy that the rate of retrograde axonal transport in a given species of

284 neurons is the same for NGF, tetanus toxin 2~, DBH-antibodies 11 and cholera toxin. amounting to about 3 mm/h for the adrenergic, 7.5 mm/h for the motor and 13 mmJh for the sensory neurons. This suggests that these macromolecules are transported by a common mechanism within a given neuron. For NGF, tetanus toxin and in the central nervous system for horseradish peroxidase it has been shown by ultrahistochemical and electron microscopic autoradiographic studies that the transported material in the axons in mainly present is vesicles and cisternae of the smooth endoplasmic reticulum18,21, 22. This suggests that the retrograde axonal transport is also linked to a system of channels of endoplasmic reticulum as postulated by Droz and collaborators for the orthograde axonal transport 9. It remains to be established whether the macromolecules are transported within these channels, or whether the membranes delineating this tubular system move as a complex together with the macromolecules in question. Moreover, the function and the localization o f the motility system which drives the retrograde axonal transport remain also to be elucidated. In contrast to cholera toxin, the retrograde axonat transport of tetanus toxin could only be blocked incompletely by both a crude mixture o f gangliosides and b y t h e purified trisialoganglioside GTI. The incomplete blockade of tetanus toxin transport could be explained by a second binding site for tetanus toxin at the nerve terminal membrane which is not a ganglioside. In fact, two populations of receptors for tetanus toxin on neuroblastoma cells have been demonstrated recently 28. Only one of the two binding sites is sensitive to neuraminidase treatment. The absence of blockade of retrograde transport of N G F and wheat germ aggtutinin by gangliosides supports the concept of specific binding sites for macromolecules on the nerve terminal membrane as a prerequisite for a highly efficient retrograde axonal transport. The receptive sites for wheat germ agglutinin probably glycoproteins are present m all neurons studied so far. whereas the specific binding sites for N G F are restricted to the adrenergic and sensory neurons. ACKNOWLEDGEMENT This work was supported by the Swiss National Foundation for Scientific Research (Grant 3.432.74).

REFERENCES l Bizzini,B., Turpin, A. and Reynand, M., On the structure of tetanus toxin, Naunyn-Schmiedebergs Arch. exp. Path. Pharrnak. 276 (1973) 271-288. 2 Bloch, R. and Burger, M. M., Purification of wheat germ agglutinin using affinitychromatography on chitin, Biochem. biophys. Res. Commun., 58 (1974) 13-19. 3 Bocchini, V. and Angeletti, P. U.. The nerve growth factor: purification as a 30,000 molecular weight protein, Proc. nat. Acad. Sci. ( Wash. ), 64 (1969) 787-794. 4 Cuatrecasas. P., Gangliosides and membrane receptors for cholera toxin, Biochemistry. 12 t 1973) 3558-3566. 5 Cuatrecasas, P., Bennett, V., Craig. S., O'Keefe, E. and Sahyoun. N., Cholera toxin, membrane glycolipids and the mechanism of action of peptide hormones. In Y. Hafefi and L. Djavadi-Ohaniance (Eds.), The Structural Basis o f Membrane Function. Academic Press, 1976. 6 Cuatrecasas, P., Parikh. I. and Hollenberg, M. D.. Affinitychromatography and struclural analys~s

285

7

8 9

10 I1

12 13 14 15 16 17 18

19 20

21

22

23

24 25

26

27 28

of vibrio cholera enterotoxin-ganglioside agrose and the biological effects of ganglioside containing soluble polymers, Biachemistry, 12 (I 973) 4253-4264. Dawson, G., Kemp, S. F., Stoolmiller, A. C. and Dorfman, A., Biosynthesis of g[ycosphingolipids by mouse neuroblastoma (NB41A), rat gila (RGC-6) and human gila (CHB-4) in cell culture, Biochem. biophys. Res. Commun., 44 (1971) 687 694. Derry, D. M. and Wolfe, L. S., Gangliosides in isolated neurons and gila[ cells, Science, 158 (1967) 1450-1452. Droz, B., Rambourg, A. and Koenig, H. L., The smooth endoplasmic reticulum: structure and role in the renewal of axonal membrane and synaptic vesicles by fast axonal transport, Brain Researeh, 93 (1975) 1-13. Fenton, E. L., Tissue culture assay of nerve growth factor and of the specific antiserum, E~-p. Cell Res., 59 (1970) 383 358. Fillenz, M., Gagnon, C., Stoeckel, K. and Thoenen, H., Selective uptake and retrograde axonal transport of dopamine {]-hydroxylase antibodies in peripheral adrenergic neurons. Brain Research 114 (1976) 293-303. Gill, D. M., Protein toxins that act within cells, Bull. d'hlst. Pasteur, 74 (1976) 65 84. Hendry, I. A., Stoeckel, K., Tboenen, H. and lversen, L. L., Retrograde axonal transport of nerve growth factor, Braht Research, 68 (1974) 103-121. Heyningen, S. van, Cholera toxin: interaction of subunits with ganglioside Gs~, Science, 183 (1973) 656 657. Heyningen, W. E. van, Gangliosides as membrane receptors for tetanus toxin, cholera toxin and serotonin, Nature (Lond.), 249 (1974) 415-417. Hol[enberg, M. D., Fishman, P. H., Bennett, V. and Cuatrecasas, P., Cholera toxin and cell growth: Role of membrane gangliosides, Proc. nat. Acad. Sei. (Wash.), 71 (1974) 4224 4228. Klein, F. and Mandel, P., Gangliosides of the peripheral nervous system of the rat, Life Sci., 16 (1975)751 758. Nauta, H. J. W., Kaiserman-Abramof, I. R., and Lasek, R. J., Electron microscopic observations of horseradish peroxidase transported from the caudoputamen to the substantia nigra in the rat: possible involvement of the agranular reticulum, Brain Research, 85 (1975) 373 384. Paravicini, U., Stoeckel, K. and Thoenen, H., Biological importance of retrograde axonal transport of nerve growth factor in ad renergic neurons, Brabl Research, 84 (1975) 279- 29 I. Schwab, M., Agid, Y., Glowinski, J. and Thoenen, H., Retrograde axonal transport of ~e'~l-tetanus toxin as a tool for tracing fiber connections in the central nervous system : connections of the rostral part of the rat neostriatum, Brain Research, (1976) in press. Schwab, M. and Tboenen, H., Selective trans-synaptic migration of tetanus toxin after retrogradc axonal transport in peripheral sympathetic nerves: A comparison with nerve growth factor, Brain Research, 122 (1976) 459-474. Schwab, M. E. and Thoenen, H., Electron microscopic autoradiographic and cytochemical localization of retrogradely transported nerve growth factor (NGF) in the rat sympathetic ganglion, J. Cell Biol., 70 (1976) 289. Stoeckel, K., Guroff, G., Schwab, M. E. and Thoenen, H., The significance of retrograde axonal transport for the accumulation of systemically administered nerve growth factor (NG F) in the rat superior cervical ganglion, Braht Research, 109 (I 976) 271 -284. Stoeckel, K., Paravicini, U. and Thoenen, H., Specificity of the retrograde axonal transport of nerve growth factor, Brah7 Research, 76 (1974)413 -421. Stoeckel, K., Schwab, M. and Thoenen, H., Specificity of retrograde transport of nerve growth factor (NGE) in sensory neurons: A biochemical and morphological study, Brahl Research, 89 (1975) 1 14. Stoeckel, K., Schwab, M. and Thoenen, H., Comparison between the retrograde axonal transport of nerve growth factor and tetanus toxin in motor, sensory and adrenergic neurons, Braht Research, 99 (1975) I 16. Svennerholm, L., Quantitative estimation of sialic acids. 11. A colorimetric resorcinol hydroch[oric acid method. Biochim. biophys. Acta (Amst)., 24 (I 957) 604-611. Zimmerman, J. M. and Piffaretti, J. CI., Interaction of tetanus toxin and toxoid with cultured neuroblastoma cells : analysis by immunofluorescence, Naunyn-Schmiedebergs Arch. exp. Path. Pharmak., 296 (1977) 271-277.