Distribution of membrane glycoproteins among the organelles of a single identified neuron of aplysia. I. association of a [3H]glycoprotein with vesicles

Brain Research, 207 (1981) 17-32 © Elsevier/North-Holland Biomedical Press

17

D I S T R I B U T I O N OF M E M B R A N E G L Y C O P R O T E I N S A M O N G T H E O R G A N E L L E S OF A S I N G L E I D E N T I F I E D N E U R O N OF APLYSIA. I. ASSOCIATION O F A [3H]GLYCOPROTEIN W I T H VESICLES

RICHARD T. AMBRON, ARIEL A. SHERBANY, LUDMILA J. SHKOLNIK and JAMES H. SCHWARTZ Departments of Anatomy, Physiology and Neurology, Division of Neurobiology and Behavior, College of Physicians and Surgeons, Columbia University, New York, N. Y. 10032 (U.S.A.)

(Accepted July 17th, 1980) Key words: membrane glycoprotein -- vesicles -- intracellular microinjection -- identified Aplysia

neurons

SUMMARY [aH]N-acetylgalactosamine injected into the cell body of R2, the giant cholinergic neuron in the abdominal ganglion, is rapidly incorporated into membrane glycoprotein and glycolipid. Incorporation, which is localized to the injected cells, occurs at a constant rate for approximately 15 h. By that time, 83 ~ of the labeled macromolecules are associated with membranes. Quantitative electron microscopic radioautography of the cell body shows that labeling of membranous organelles is selective: the Golgi apparatus, endoplasmic reticulum, and lucent and compound vesicles are labeled, while the nucleus, end-stage lysosomes, and mitochondria are not. SDS-polyacrylamide gel electrophoresis of the total membranes from more than 40 R2s examined individually resolves reproducibly 5 major labeled glycoprotein components. In order to determine which of these are associated with vesicles, we isolated a labeled vesicle fraction from R2 using a combination of differential centrifugation and filtration on a column of glass beads with 200 nm pores. This fraction was consistently enriched in [aH]glycoproteins I (180,000 Daltons) and V (90,000 Daltons) relative to those fractions containing larger organelles. These experiments suggest that different organelles contain characteristic membrane components.

INTRODUCTION Membranes of various intracellular organelles differ in macromolecular composition ~9. In the neuron, where each region - - cell body, axon, terminal - - is distinct

18 morphologically and functionally, there are regional differences in surface membranes as well. These differences seem to be reflected in glycoprotein composition. For example, specific lectin binding sites are localized at synaptic terminals 9,25,2s and the Na+-K+-ATPase 10 and acetylcholine receptor 3~ are regionally distributed. Since most if not all macromolecules are synthesized in the cell body, selective mechanisms must exist for distributing specific membrane components to organelles and for segregating them regionally. Studies on axonal transport have provided some insight into these processes4,1~,~3,~7,a6. Certain glycoproteins must remain in the cell body, however, in order to differentiate this region of the neuron from its processes 3,4. We have been examining the synthesis and axonal transport of membrane components in single identified neurons of Aplysia 1-3. These neurons are large enough to allow direct intrasomatic injection of labeled precursors, and have an anatomical arrangement that readily permits separation of cell body from axon for experimental analysis. In our earlier studies on the giant neuron R2 we found that about 35 ~ of the newly synthesized glycoprotein labeled after intrasomatic injection of [ZH]fucose is incorporated into a glycoprotein with a molecular weight of 180,000 daltons, which we call Component-IZ, 3. This glycoprotein is preferentially retained in the cell body, whereas other, lower molecular weight glycoproteins are preferentially exported into the axona,3, 4. In this and the accompanying paper we present evidence using quantitative radioautography and subcellular fractionation that Component-I is a constituent of a specific type of vesicle in R2's cell body. MATERIALS AND METHODS

Aplysia californica, weighing 50-150 g, were supplied by Pacific Bio-Marine, Venice, Calif., and maintained at 15 °C 1,12. lntrasomatic injection L-[1,5,6-3H]fucose (4.8 Ci/mMol) or [G-ZH]N-acetyl-D-galactosamine (25 Ci/mmol) (New England Nuclear, Boston, Mass.) were prepared and injected into R23,12. Radioactivity was counted by liquid scintillation using purified Triton X-100 (New England Nuclear). Under our conditions of counting, one pmole of [3H]Nacetylgalactosamine corresponded to 10,500 cpm. After injection, the nervous system was maintained at 15 °C in artificial seawater supplemented with amino acids and vitamins 12. In experiments requiring periods of maintenance 10 h or longer, R2 was impaled again prior to fractionation to ensure that the cell was in good condition. Not all of the injected radioactivity remained within R2. At one hour after injection 13 -4- 8 ~ (n -- 3)* of the total injected radioactivity could be found in the bath. By 5 h the value was 39 -k- 7 ~ (n = 8); it remained constant thereafter, indicating that essentially all of the radioactivity which was going to escape from the cell had done so by 5 h. General uptake of [aH]N-acetylgalactosamine from the bath is inefficient6 and labeling of other neurons, glial cells and connective tissue is negligible (see Results). * All values are expressed as mean q- S.E.M.

19

Fractionation of the injected neuron Details of the fractionation procedures have been described3, 6,a7. Since injected sugar precursors are incorporated only within the injected cell, and since labeled macromolecules remain localized only to that cell and its processes (see Results), it was not necessary to remove the cell body or axon from the surrounding tissues for biochemical analyses. The washed abdominal ganglion containing an injected R2 cell body, and the right connective, containing R2's major axon, were homogenized, centrifuged at 105,000 × g for 45 min, and the pellet washed 3 times with buffer. The combined supernatant was treated with 5 ~ trichloroacetic acid (TCA)-0.5 ~ phosphotungstic acid (PTA) and the precipitate, containing soluble glycoprotein, was collected on a glass fiber pad (Whatman GF/C) for counting by liquid scintillation. The tissue pellet was sequentially extracted at 4 °C with 30 vols. of distilled chloroform-methanol 2:1 and 1:2 (v/v). After [3H]glycolipids had been removed 87, [3H]glycoproteins were extracted from the pellet with sodium dodecyl sulfate (SDS)-mercaptoethanol at 70 °C, and the small amount of residual glycoprotein was solubilized with 90 ~ aqueous formic acid. All of the radioactivity in R2 is recovered by this procedure z. The sum of the radioactivity in the SDS and formic acid extracts was taken as the particulate glycoprotein content of the neuron.

Dissection of R2's cell body and subcellular fractionation of the cytoplasm The abdominal ganglion containing an injected R2 was pinned to silicone plastic (Sylgard, Dow Chemical, Midland, Mich.) and the connective tissue sheath just above R2's cell body was removed. The cell body, teased free of its axon, was transferred to a 50/~1 drop of artificial seawater maintained at 4 °C. A single opening was made in the membrane of the isolated cell. If done carefully, the nucleus emerges intact followed by the cytoplasmic contents of the cell 27. Nucleus, cell envelope, and cytoplasm were collected separately with micropipettes. Labeled cytoplasm collected from the dissected cell body was added to a lowspeed supernatant (1000 × g, l0 min) from unlabeled Aplysia nervous tissue which had been homogenized in a Potter-Elvejhem tissue grinder (clearance 0.25 mm; Kontes Glass Vineland, N.J.) containing isotonic buffer (0.3 M NaC1, 0.2 M sucrose, 1 mM EDTA, 10 mM Tris.HCl, pH 8.0). The low-speed supernatant, now containing R2's cytoplasm, was centrifuged at 15,000 × g for 5 min in a modification of the procedure of Johnson and Lardy 23 originally designed to separate mitochondria and other large organelles from microsomes. The pellet was resuspended in the buffer by 5 up and down strokes of a loose-fitting homogenizer. Centrifugation of the suspension at 600 × g for 5 min yielded a crude pellet (Pc) and a supernatant that was then centrifuged at 15,000 × g for 5 min, to give the mitochondrial pellet (PM). The supernatant was combined with the initial 15,000 × g supernatant to yield ($15) for application to a column of glass beads with a nominal pore diameter of 202 ~ 11.5 nm (CPG 2000, 80/120 mesh, Electro-nucleonics, Fairfield, N.J.) at 4 °C. All solutions were de-aerated prior to passage through the column. Before each use, the beads in the column were treated with a 1 ~ solution of polyethylene glycol 20-M (Union Carbide) to minimize tissue adsorption 19 and were then equilibrated with homogenizing buffer. The

20 exclusion volume was determined using polystyrene spheres with a nominal mean diameter of 312 nm (Ernest F. Fullam, Schenectady, N.Y.). Polystyrene was detected in eluant fractions by measuring Ae59 in dioxane. Colloidal gold particles with mean diameters of 10, 60 and 90 nm were used to test the resolution of the column. The particles were prepared by reduction of gold chloride with sodium citrate 14 and their diameters measured by electron microscopy. The totally included volume of the column was determined with [3H]N-acetylgalactosamine. Discontinuous density gradients. The vesicle fraction from the glass bead column was layered onto an iso-osmolar discontinuous gradient with density steps of 0.4 M sucrose, 0.7 M sucrose, and 15 ~ (w/v) Ficoll (Pharmacia, Piscataway, N.J.) in 0.8 M sucrose (equivalent in density to 1.2 M sucrose). Gradients were centrifuged for 90 min at 81,000 × g average in an SW 27 rotor (Beckman Instruments, Spinco Div., Palo Alto, Calif.). Fractions were collected from the top using an Autodensiflow (Buchler Instruments, Fort Lee, N.J.).

Preparation of tissue for microscopy and radioautography The abdominal ganglion containing an injected R2, or pellets obtained from subcellular fractionation experiments, were prepared for microscopy and radioautography as previously described 38. Sections were examined in a Philips EM 300. Background over areas not containing tissue was always negligible. Quantitative radioautography. The distribution of silver grains on radioautographs enlarged 30,1300 times was analyzed by the method of Williams 4°. A circle was drawn around each grain, centered on the midpoint of its longest axis. The circle had a radius 1.5 times the half-distance (the distance from a radioactive line within which 50 ~ of the developed grains fall). For the conditions used (Ilford L4 emulsion and development in Kodak Microdol-X) Salpeter et al. 83 have determined the halfdistance to be 165 nm. Potential isotope sources within the circle were scored as single, junctional, or compound items 40. To determine the relative volumes occupied by the various organelles, the same micrographs were covered with a regular grid of circles identical in size to those used for scoring grains. The frequency with which a cellular component appeared within the circles was recorded and was taken as its effective area. Grains in junctional and compound categories were apportioned to the individual organelles by point-counting. Circle and grain frequency distributions were compared by a Z2 test to determine whether the grain distribution differed significantly from random. To ensure that grains over radioautographs were associated with macromolecules, we measured radioactivity appearing in the fixative and other solutions used to prepare the tissue for microscopy. Two R2s were injected with [ZH]Nacetylgalactosamine; 5 h after injection, the tissues were fixed separately and processed for microscopy. Results from both cells agreed closely. Of the total radioactivity in the two cells, 62 and 65 ~ appeared in the wash solutions during fixation. These values are in good agreement with the proportion of radioactivity in the cell (69 ~o at 5 h) that was found to be acid-soluble in biochemical fractionation studies (see Results). Of the radioactivity that was washed out during fixation, 88 ~ was found in the glutaralde-

21 hyde, 6 ~ in the collidine buffer, none in osmium, 5 ~ in ethanol and 1 ~ in propylene oxide. It is likely that glycolipids are removed from the tissue by this fixation protocol ~4. Polyacrylamide gel electrophoresis. Labeled membrane glycoproteins were examined by electrophoresis in SDS on 5 ~ continuous polyacrylamide gels as described 1, ag, or by the discontinuous method of Laemmli 26. When Laemmli's method was used, particulate fractions were extracted for 20 min with 2 ~ S D S - 5 ~ 2-mercaptoethanol in 0.063 M Tris.HCl (pH 7.1) at 70 °C. Insoluble material was removed by centrifugation and the extract electrophoresed in SDS on a 3 ~ stacking gel-7.5 running gel. Dansylated beef serum albumin was included in each sample as an internal standard for estimating molecular weight 21. The procedures for cutting gels and for counting radioactivity in the slices have been described1, 6. The per cent radioactivity in individual glycoproteins was estimated on graph paper by plotting radioactivity (cpm) against distance along the length of the gel. The proportion of the radioactivity associated with the glycoproteins was measured by cutting out and weighing the paper under the peaks. In other experiments, the particulate fraction was extracted with SDS, reduced with dithiothreitol, and alkylated with N-ethyl maleimidO s. Samples were then electrophoresed on a 10 ~ polyacrylamide slab with a 3 ~ stacking gel. The slab was dried and developed for radioautography s. RESULTS

Distribution of [3H] N-acetylgalactosamine after intrasomatic injection [3H]N-acetylgalactosamine injected into the cell body of R2 was incorporated primarily into membrane macromolecules. We first examined the distribution of incorporated radioactivity within a ganglion containing an injected R2 by light microscopic radioautography. Silver grains were localized only over the injected cell; there was essentially no labeling of glial cells, connective tissue, or neighboring neurons. Within R2, most of the grains were located over cytoplasm with only a few grains over the nucleus (Fig. 1). This restricted distribution is similar to that previously observed with [3H]fucose3,Ss. By one hour after injection, radioactive glycoprotein and glycolipid began to appear in R2's major axon that runs within the right connective. A detailed analysis of the export of these labeled components from the cell body and their rapid translocation along the axon will be presented elsewhere. Statistical analysis of electron microscopic radioautographs of R2's cell body 15 h after injection showed that silver grains were not distributed randomly, but were associated with only certain organelles (Table I). Both compound and electron-lucent vesicles were significantly labeled. Compound vesicles consist of a small vesicle enclosed within a larger one (Fig. 2 and refs. 6, 16, 38). Because of the small size of these organelles, the extent of their labeling is considerably underestimated 34. Also labeled were the Golgi apparatus, where vesicles form20, 29, and the endoplasmic reticulum, where nascent glycoproteins are glycosylated7. Incorporation of [3H]Nacetylgalactosamine into macromolecules begins to diminish only at 15 h after

22

Fig. 1. Light microscopic radioautograph of a 2 ffm-transverse section of abdominal ganglion 15 h after i ntrasomatic injection of R2 with [3H]N-acetylgalactosamine. Silver grains are located almost exclusively over R2's cytoplasm (Cyt) with negligible labeling of nucleus (Nu), surrounding glial cells (G1), connective tissue sheath (Sh), or other neurons (N). × 91.3. Fig. 2. Electron microscopic radioautograph with R2's cell body showing organelles labeled 15 h after injection (see Table I): Golgi apparatus, G; lucent vesicle, LV; compound vesicle, CV; and multivesicular body, MVB. Not significantly labeled are mitochondria (M). Radioautograph was exposed for 5 days after injection of a 50 g animal with [aH]N-acetylgalactosamine. x 24,900.

23 TABLE I Distribution of silver grains over R2"s cell body 15 h after injection of [ZH] N-acetylgalactosamine

For the analysis, 381 silver grains and 1309 area circles were counted. Grain distribution (%)

Labeled* Golgi membranes Rough endoplasmic reticulum Multivesicular bodies Smooth endoplasmic reticulum Compound vesicles Lucent vesicles External membrane Not labeled** Nucleus Lysosomes (lipochondria) Mitochondria Cytosol Outside (Glial cells, connective tissue)

Effective area (%)

Observed Expected grains grains

Relative specific activity* **

9.2 12.4 3.6 31.8 3.2 2.5 4.9

1.0 4.5 0.8 9.9 1.0 1.1 2.2

35 47 14 121 12 10 19

3.8 17.1 3.1 37.6 3.8 4.2 8.4

9.2 2.8 4.4 3.2 3.2 2.3 2.2

0.5 4.5 2.9 21.3

1.0 6.9 2.2 58.2

2 17 11 81

8.4 26.0 8.4 221.2

0.5 0.7 1.3 0.4

2.9

10.9

11

41.4

0.3

* Items labeled to a greater extent than expected for a random distribution of grains. Significance (P < 0.05) was determined by a X2 test. ** Items labeled to a lesser extent than expected from a random distribution of grains (P < 0.05). * * * Relative spec. act. is the ratio of the frequency of grains to the effective area.

injection (see below) so it was n o t surprising to find t h a t these p r e c u r s o r organelles were labeled. The p l a s m a m e m b r a n e a n d multivesicular bodies also c o n t a i n e d s o m e r a d i o a c t i v i t y , b u t the d i s t r i b u t i o n o f grains over large lysosomes (lipochondria), m i t o c h o n d r i a , a n d the nucleus showed t h a t they h a d fewer grains t h a n w o u l d be expected f r o m a r a n d o m distribution. A s a n t i c i p a t e d f r o m inspection o f light m i c r o s c o p i c r a d i o a u t o g r a p h s (Fig. 1), glial cells, either on R2's surface o r invaginating d e e p l y into R 2 ' s c y t o p l a s m , were n o t labeled. Biochemical characterization o f [3 H]macromolecules Fractionation studies. A t all p e r i o d s after injection, m o s t i n c o r p o r a t e d r a d i o activity was associated with m e m b r a n e s (Table II). As with [3H]fucose 3, the extent o f i n c o r p o r a t i o n into m a c r o m o l e c u l e s d e p e n d e d on the a m o u n t o f labeled p r e c u r s o r present in the cell. I n c o r p o r a t i o n was linear with a m o u n t s up to a b o u t 50 p m o l ; b e y o n d that, i n c o r p o r a t i o n b e g a n to diminish. This is consistent with a recent i n d i c a t i o n t h a t soluble p r e c u r s o r s regulate g l y c o p r o t e i n synthesis 32. A l l cells used for analysis were injected with a m o u n t s o f [3H]N-acetylgalactosamine t h a t resulted in i n t r a c e l l u l a r c o n c e n t r a t i o n s o f r a d i o a c t i v e p r e c u r s o r less t h a n 50 p m o l . Biochemical analyses showed labeled m a c r o m o l e c u l e s to consist p r i m a r i l y o f m e m b r a n e g l y c o p r o t e i n s a n d glycolipids, b u t some soluble g l y c o p r o t e i n was also

24 TABLE II Distribution of [3HI N-Acetylgalaetosamine in R2 after injection The abdominal ganglion, containing R2's cell body, and the attached right connective, containing R2's axon, were homogenized and fractionated at 105,000 × g as described in Methods. Radioactivity in membrane glycoprotein is the sum of radioactivity in the SDS and formic acid extracts. Incorporation into macromolecules ( % of total radioactivity)

Soluble glycoprotein Membrane fraction Glycoprotein Glycolipid Total macromolecules Total radioactivity (pmol/cell)

lh(n=4)

5h(n=8)

l O h ( n = 5 ) 1 5 h ( n = 7) 22h(n=7)

1.8 ± 0.5

4.9 -4- 0.5

5.2 ± 0.3

1.8 ± 0.7 1.2 4- 0.1 4.8 4- 1.2 12-50

19.0 4- 2.8 7.3 4- 0.9 31.2 4- 3.8 7-23

26.1 ± 2.0 7.4 4- 0.9 38.7 ± 2.0 5-36

9.5 -- 1.8

7.9 4- 1.5

35.3 4- 4.1 40.2 4- 2.8 11.3 ± 2.5 13.5 4- 2.1 56.1 4- 3.8 61.6 4- 2.6 4-45 0.5-38

Values are mean ± S.E.M. n = number of experiments. present (Table II). Incorporation of [3H]N-acetylgalactosamine into glycoprotein begins to diminish by 15 h after injection. Gel electrophoresis studies. Examination of SDS gel electropherograms from more than 40 injected R2s by counting radioactivity in m m slices (Fig. 3A, B) or by fluorography (not shown) enabled us to identify reproducibly 5 major glycoprotein components with molecular weights greater than 55,000 daltons. Other labeled components are present but these consistently contained only a small proportion of the total radioactivity. Reduction and alkylation of the glycoproteins prior to electrophoresis did not affect the number of major components resolved nor did any new peaks appear when samples were electrophoresed on 7.5 or 10 ~ acrylamide gels. Of the 5 components, I, III, and V have mobilities similar to those of glycoproteins that were labeled when R2 was injected with [3H]fucose (compare 3A and B; see ref. 3). Thus, it is likely that these are glycoproteins that contain both fucose and N-acetylgalactosamine moieties. Since only Component-I has been isolated and characterized4, 5 however, we cannot be certain that the other components are individual glycoproteins. [3H]glycoproteins with molecular weights less than 55,000 Daltons were much more variable in their mobilities and also in the proportion of the radioactivity that they contained. Glycoproteins associated with vesicles The observation that vesicles in the cell body were labeled significantly by 15 h after injection (Table I) prompted us to examine whether any of the components could be assigned to these organelles. Our approach was to utilize a combination of techniques that would separate vesicles by size and density. As starting material we used cytoplasm obtained by dissection of isolated cell bodies (see Methods) 24-27 h after injection. Labeled cytoplasm from the injected R2 was added to a low speed supernatant prepared from unlabeled nervous tissue, and large organelles were removed by differential centrifugation. The supernatant ($15) was then filtered on a

25 D

M

I

I

A

240

'~ 180 r20 6O

I-.0

I

I

I

I

I

I

1500 f 1200 900 600 500

J /

Origin

10 20 250 40 50 60 DISTANCE (cm)

Fig. 3. Continuous SDS polyacrylamide gel electrophoresis of particulate glycoproteins from R2's cell body 15 h after intrasomatic injection of [aH]fucose(A) or [3H]N-acetylgalactosamine(B). The 105,000 × g membrane pellet was treated with chloroform-methanol and then the [3H]glycoproteins were dissolved and reduced in sample buffer. The extract was run on a continuous 7.5 ~ polyacrylamide gel. The gel was sectioned into mm segments and the radioactivity in each segment counted as described in Methods. M, monomer; D, dimer; T, trimer (of beef serum albumin).

column of glass beads that had been standardized using polystyrene beads and colloidal gold particles (see Methods). Radioactivity emerged from the column in two fractions (Fig. 4). The totally-included fraction contained soluble and low molecular weight precursors. In contrast, the partially-included radioactivity appeared with the 90 nm gold particle marker and was mostly in membranes: approximately 75 ~ could be sedimented by centrifugation and remained so, even after repeated washing with 50 m M Tris.HCl ( p H 7.6). These conditions would be expected to release trapped soluble macromolecules 3. When the membrane fraction from the column was examined by electron microscopy, it was seen to consist of membranous profiles 40-200 n m in diameter, including vesicles of varying size (Fig. 5c). Mitochondria and large lysosomes, which were abundant in PM (Fig. 5a), were few, and ribosomes, which were abundant in supernatant $15 (Fig. 5b) were also rare. It should be noted that the size range of the profiles in this fraction serves only as an internal standard for the possible sizes of labeled organelles from R2 since essentially all of the membrane profiles seen are derived from the carrier tissue which was heterogeneous. In order to examine the behavior on the column of a known vesicle population, we injected the giant cerebral neuron ( G C N ) of Aplysia with [3H]serotonin. Evidence

26 36,000 PSB

90 60 I0

BO

30,000

24,000 "o E >I-I 8,000 I--

8 r~

12,000

6,000

0

.

._ .

.

6

.

_

_ ~ " )

12

I

18

I

I

24

30

L . . . .

36

42

VOLUME (ml) Fig. 4. Filtration of supernatant $15 from cytoplasm of R2 on a column of glass beads. A typical result from one of five individual experiments is shown. The combined cytoplasm of 3 R2s (125,000-330,000 cpm/cell) was obtained by dissection 24-27 h after injection of 13H]N-acetylgalactosamine. The labeled cytoplasm, combined with the low speed supernatant from a homogenate of 50 Aplysia ganglia, was subjected to differential centrifugation and the mitochondrial supernatant ($15) was filtered on a column of glass beads (0.9 × 55 cm) prepared and standardized as described in Methods. Fractions of 1.3 ml were collected and a tithe was removed for counting. The column was calibrated with polystyrene beads (PSB), 90, 60, and 10 nm colloidal gold particles (arrows), [3H]serotonin-containing vesicles (SV) and blue dextran (BD) (Pharmacia). Material in the fractions circled was collected for examination by electron microscopy (Fig. 5), electrophoresis (Fig. 7), or was subjected to sucrose density centrifugation (Fig. 6).

has been presented that serotonin in the axons of this neuron is contained in vesicles15,16:15 h after injection, the lip nerve and cerebrobuccal connective that contain the axons of GCN were homogenized and fractionated. When supernatant $15 was applied to the glass bead column, labeled vesicles emerged as a single peak whose position corresponded to that of the 90 nm gold particles and the membranes from R2 (Fig. 4).

Fig. 5. Examination of subcellular fractions by electron microscopy. Samples were fixed and prepared for electron microscopy as described in Methods. All figures are at same magnification: 24,900 ×. a : contents of pellet, PM. Many large lysosomes (L) and mitochondria (M) are seen, and vesicles of various types are also present, b: organelles in supernatant $1~after isolation by centrifugation at 105,000 × g for 1 h. Fraction consists of vesicles interspersed among ribosomes, c: fraction from glass bead column after centrifugation at 105,000 × g. Many vesicles and smooth membrane profiles of irregular shape are present. There are few ribosomes, and mitochondria are rare. d : composition of membrane fraction above the 0.7 M interface from iso-osmolar sucrose density gradient centrifugation (see text). The band was collected from the gradient, diluted 3-fold with artificial seawater, and then centrifuged at 105,000 × g to isolate the membranes. The field contains primarily small vesicles. Lucent and compound vesicles are present as are abundant irregular profiles of unknown origin.

t , JP "-.,.I

28 ,02M tQI4M,Q7M , 12 M,

"~1 sv I~

900

B

Q. (.}

>t-.> }.-

600

0

0

' C 300

01~5

15

~ 25

I

35

Top

FRACTION

NUMBER

Fig. 6. Density gradient centrifugation of the membrane fraction from glass bead column. Fractions circled in Fig. 4 were combined, layered on a 3-step iso-osmolar discontinuous sucrose-Ficoll gradient,

and centrifugedat 81,000 × g for 90 rain. 0.65 ml fractionswere collectedand sampleswere removed for counting. Vesicles from the column were further characterized by density gradient centrifugation. The distribution of vesicles containing 3H-labeled membrane glycoproteins was compared with that of vesicles labeled with serotonin. Three bands containing labeled glycoprotein were detected (Fig. 6). Two major bands (A and B, Fig. 6) appeared in the region of the gradient which corresponded to that observed with the serotonergic vesicles. A third band (C, Fig. 6) contained a smaller proportion of the radioactivity, and was found at the 1.2 M interphase. Visual examination of the gradient showed most membranes to be on the 1.2 M interphase. A much smaller amount was seen above the 0.7 M interface which did, however, contain most of the radioactivity. Examination of the material from the gradient by electron microscopy showed it to consist of compound vesicles and other vesicular profiles (Fig. 5d). When these procedures were carried out using cytoplasm from an R2 injected with [3H]fucose, the distribution of labeled membrane on the column and gradient was the same as that obtained with [3H]N-acetylgalactosamine. Fucose also labels compound and lucent vesicles3s and since fucose incorporation attains a maximum by 10 h after injection3, at 27 h we would expect a considerable proportion of labeled cytoplasmic membranes to be in vesicles.

Characterization of partially purified vesicle membrane [3H]glycoproteins by gel electrophoresis To ascertain the distribution of cytoplasmic N-acetylgalactosamine-labeled glycoproteins among the various fractions, we examined by electrophoresis the composition of membranes from pellets Pc and PM, and the membranes from the

29 I T

75O

M

m

A

600 45O 5OO 150 0 1200 I

B

I000 I.-I--

800! 600 400

121

200

6o8 5OO 4OO 300

~

IC

2OO I00

20

40

60

DISTANCE (mm)

Fig. 7. Discontinuous polyacrylamidegel electrophoresisin SDS of [ZH]glycoproteinsin membrane fractions obtained by subcellular fractionation. Lipids were removed prior to electrophoresis. A: membrane fraction from glass bead column. B: pellet P~t. C: pellet Pc. column., Three experiments were carried out, each using the combined cytoplasm from 3 or more R2 cell bodies 24 h after injection. In each experiment, two of the [3H]glycoprotein components, I and V, were found to be enriched in the vesicle fraction from the column, while the intermediate molecular weight glycoproteins predominated in pellets Pc and P~t (Fig. 7). On the average, 53 q- 6 % of the total Component-I, and 55 4- 13 % of the Component-V in the cytoplasm was in the vesicle fraction from the column. We subjected this fraction to the density gradient centrifugation already described (Fig. 6) but were unable to separate organelles containing a single component. Both Components I and V were present in bands A and B in approximately the same proportion, despite the difference in density of the vesicles. Several additional labeled components, not seen in the other fractions, were found associated with membranes in pellet Pc (Fig. 7C). Electron microscopy has revealed that this fraction contains multivesicular bodies and other lysosomal structures; it is therefore possible that these glycoproteins are degradation products.

30 DISCUSSION A major objective of our research is to associate unique, identified [3H]glycoproteins with specific organelles: in this way, the fate of these organelles in the various regions of the neuron can be followed. We chose to work with R2 because its large size confers several experimental advantages. Aside from the ease with which the cell can be injected, the cell body is readily removed from the ganglion and the cytoplasm obtained by dissection (see Methods). Using this technique, cytoplasmic membranes are exposed to a minimum of disruption. We have assessed the efficacy of the dissection by assaying the distribution of the soluble enzyme choline acetyltransferase: approximately 95 ~ of the total cytoplasm is isolated by this procedure (see refs. 4, 5 and Ambron and Schwartz, manuscript in preparation). When R2 is injected with [3H]fucose or [3H]N-acetylgalactosamine much of the radioactivity is incorporated into membrane glycoproteins (Table II) that are associated with only certain organelles, including vesicles (Table I). When the distribution of labeled membrane glycoproteins was examined by a variety of subcellular fractionation techniques, we found two 3H-labeled components to be enriched in vesicle fractions. The other components were associated primarily with larger organelles (Fig. 7). One of these, [aH]glycoprotein-I (I 80,000 daltons) has been isolated and shown to contain [3H]fucose and [3H]N-acetylgalactosaminea,5. The other, Component-V (90,000 daltons), has not been as well characterized, but also appears to be labeled with both sugars. [3H]Component-I and Component V fractionated with vesicles, as indicated both by the behavior of vesicles in the unlabeled carrier tissue and of the [3H]serotonin-labeled vesicles. We therefore infer that the radioactive glycoproteins are components of organelles with the characteristics of vesicles. The question is, what is the origin of these organelles? Did they already exist in R2's cytoplasm, or were they artifactually derived from the Golgi and endoplasmic reticulum despite the precautions we took in obtaining the cytoplasm by dissection? Since we cannot obtain enough cytoplasm from R2 to examine the fractions by microscopy or to assay marker enzymes, we cannot assess the magnitude of contamination by Golgi and reticulum directly. We do known however, that by 15 h after injection at least 1 0 ~ of the membrane-associated silver grains are over compound and lucent vesicles. By 24-27 h, this percentage is likely to be higher since incorporation diminishes after 15 h (Table II). It is unlikely that all of the radioactivity in the vesicle fraction is produced artifactually since compound vesicles can be readily identified in the cell and are also plentiful in the vesicle fraction from the gradient. In the accompanying paper 5 we show that the protein synthesis inhibitor, anisomycin, causes a marked shift in the distribution of labeled glycoproteins. The compound vesicle becomes the most heavily labeled organelle and evidence is presented that it contains Component-1, The association of specific glycoproteins with vesicles is important since it indicates that glycoproteins are not uniformly distributed throughout all of the organelles of a neuron.

31 ACKNOWLEDGEMENTS The a u t h o r s t h a n k Ms. Lise Castellucci for expert technical assistance, a n d Drs. M. G e r s h o n , D.J. Goldberg, J.E. G o l d m a n a n d E.R. K a n d e l for critical reading of the manuscript. This research was supported by N a t i o n a l Institutes of Health Research G r a n t s NS 12066 (to J.H.S.) a n d by NS 14555 a n d Career D e v e l o p m e n t A w a r d NS 00350 (to R.T.A.). A.A.S. was supported by P H S Medical Scientist T r a i n i n g G r a n t s GMO-1668 a n d GMP-6308 ; d u r i n g part of this work, R.T.A. was a Senior Investigator of the New Y o r k H e a r t Association.

REFERENCES 1 Ambron, R. T., Goldman, J. E. and Schwartz, J. H., Axonal transport of newly synthesized glycoproteins in a single identified neuron of Aplysia californica, J. Cell Biol., 61 (1974) 665-675. 2 Ambron, R. T., Goldman, J. E. and Schwartz, J. H., Effect of inhibitingprotein synthesis on axonal transport of membrane glycoproteins in an identified neuron of Aplysia, Brahl Research, 94 (1975) 307-323. 3 Ambron, R. T., Goldman, J. E., Thompson, E. B. and Schwartz, J. H., Synthesis of glycoproteins in a single identified neuron of Aplysia californica, J. Cell Biol., 61 (1974) 649-664. 4 Ambron, R. T. and Schwartz, J. H., Regional aspects of neuronal glycoprotein and glycolipid synthesis. In R. U. Margolis and R. K. Margolis (Eds.), Complex Carbohydrates of Nervous Tissue, Plenum Press, New York, 1979, pp. 269-290. 5 Ambron, R. T., Sherbany, A. A. and Schwartz, J. H., Distribution of membrane glycoproteins among the organelles of a single identified neuron of Aplysia. lI. Isolation and characterization of a glycoprotein associated with vesicles, Brain Research, 207 (1981) 33-48. 6 Ambron, R. T. and Treistman, S. N., Glycoproteins are modified in the axon of R2, the giant neuron of Aplysia californica, after intra-axonal injection of [3H]N-acetylgalactosamine, Brain Research, 121 (1977) 287-309. 7 Blobel, G. and Dobberstein, B., Transfer of protein across membranes. II. Reconstitution of functional rough microsomes from heterologous components, J. Cell Biol., 67 (1975) 852-862. 8 Bonnet, W. M. and Laskey, R. A., A film detection method for tritium-labeled proteins and nucleic acids in polyacrylamide gels, Europ. J. Biochem., 46 (1974) 83-88. 9 Cotman, C. W. E. and Taylor, D., Localization and characterization of concanavalin A receptors in the synaptic cleft, J. Cell Biol., 62 (1974) 236-242. 10 Dahl, J. L. and Hokin, L. E., The sodium-potassiumadenosine-triphosphatase, Ann. Rev. Biochem., 43 (1974) 327-356. 11 Droz, B., Synaptic machinery and axoplasmic transport: maintenance of neuronal connectivity. In D. B. Tower (Ed.), The Nervous System, Vol. 1, Raven Press, New York, 1975, pp. 111-127. 12 Eisenstadt, M., Goldman, J. E., Kandel, E. R., Koike, H., Koester, J. and Schwartz, J. H., Intrasomatic injection of radioactive precursors for studying transmitter synthesis in identified neurons of Aplysia californica, Proc. nat. Acad. Sci. (Wash.), 70 (1973) 3371-3375. 13 Elam, J. S., Axonal transport of complex carbohydrates. In R. U. Margolis and R. K. Margolis (Eds.), Complex Carbohydrates of Nervous Tissue, Plenum Press, New York, 1979, pp. 235-268. 14 Frenz, G., Controlled nucleation for the regulation of the particle size in monodispersed gold suspensions, Nature phys. Sci., 241 (1973) 20-22. 15 Goldberg, D. J., Goldman, J. E. and Schwartz, J. H., Alterations in amounts and rates of serotonin transported in an axon of the giant cerebral neurone of Aplysia californica, J. Physiol. (Lond.), 259 (1976) 473-490. 16 Goldman, J. E., Kim, K. S. and Schwartz, J. H., Axonal transport of [3H]serotonin in an identified neuron of Aplysia californica, J. Cell Biol., 70 (1976) 304-318. 17 Grafstein, B. and Forman, D., Intracellular transport in neurons, Physiol. Rev., 60 (1980) 11671283. 18 Hamilton, S. L., McLaughlin, M. and Karlin, A., Formation of disulfide-linked oligomers of acetylcholine receptor in membrane from Torpedo electric tissue, Biochemistry, 18 (1979) 155-163.

32 19 Hiat, C.W.,Shelokov, A.,Rosenthal, G. J. andGalimore, J. M.,Treatment ofcontrolled pore glass with poly (ethylene oxide) to prevent adsorption of rabies virus, J. Chromat., 56 (1971) 362-364. 20 Holtzman, E., The origin and fate of secretory packages, especially synaptic vesicles, Neuroscience, 2 (1977) 327-355. 21 Inouye, M., Internal standards for molecular weight determinations of proteins by polyacrylamide gel electrophoresis, J. bioL Chem., 246 (1971) 4834-4838. 22 Jamieson, J. D. and Palade, G. E., Production of secretory proteins in animal cells. In B. R. Brinkly and K. R. Porter (Eds.), International CellBiology 1967-1977, Rockefeller Univ. Press, New York, 1977, pp. 308-317. 23 Johnson, D. and Lardy, H., Isolation of liver and kidney mitochondria. In R. W. Estabrook and M. E. Pullman (Eds.), Methods in Enzymology, VoL X, Academic Press, New York, 1967, pp. 94-96. 24 Johnson, P. V. and Roots, B. I., Nerve Membranes, Pergamon Press, New York, 1972, p. 122. 25 Kelly, P., Cotman, C. W., Gentry, C. and Nicolson, G. L., Distribution and mobility of lectin receptors on synaptic membranes of identified neurons in the central nervous system, J. Cell BioL, 71 (1976) 487-496. 26 Laemmli, U. K., Cleavage of structural proteins during the assembly of the head of bacteriophage T4, Nature (Lond.), 277 (1970) 680-685. 27 Lasek, R. J. and Dower, W. J., Analysis of nuclear DNA in individual nuclei of giant neurons, Science, 172 (1971) 278-280. 28 Matus, A., DePetris, S. and Raft, M. C., Mobility ofconcanavalin A receptors in myelin and synaptic membranes, Nature New Biology, 244 (1973) 278-280. 29 Meldolesi, J., Secretory mechanism in pancreatic acinar cells. Role of cytoplasmic membranes. In B. Ceccarelli, F. Clementi and J. Meldolesi (Eds.), Advances in Cytopharmacology, Vol. 2, Raven Press, New York, 1974, pp. 71-85. 30 Nagy, A., Baker, R. R., Morris, S. J. and Whittaker, V. P., The preparation and characterization of synaptic vesicles of high purity, Brain Research, 109 (1976) 285-309. 31 Rosenberry, T. L., Chen, Y. T. and Bock, E., Structure of 11S acetylcholinesterase. Subunit composition, Biochemistry, 13 (1974) 3068-3079. 32 Richards, W. L., Kilker, R. D. and Sherif, G. S., Metabolite control of e-fucose utilization, J. biol. Chem., 253 (1978) 8359-8361. 33 Salpeter, M. M., Bachman, L. and Salpeter, E. G., Resolution in electron microscope radioautography, J. Cell Biol., 41 (1969) 1-20. 34 Salpeter, M. M. and McHenry, F. A., Electron microscope autoradiography, analysis ofautoradiograms. In J. K. Koehler (Ed.), Advanced Technique in Biological Electron Microscopy, Springer, New York, 1973, pp. 114-151. 35 Salvaterra, P. M., Gurd, J. M. and Mahler, H. R., Interactions of the nicotinic acetylcholine receptor from rat brain with lectins, J. Neurochem., 29 (1977) 345-348. 36 Schwartz, J. H., Axonal transport: components, mechanisms, and specificity, Ann. Rev. Neurosci., 2 (1979) 467-504. 37 Sherbany, A. A., Ambron, R. T. and Schwartz, J. H., Membrane glycolipids: regional synthesis and axonal transport in a single identified neuron of Aplysia californica, Science, 203 (1979) 78-81. 38 Thompson, E. B., Schwartz, J. H. and Kandel, E. R., A radioautographic analysis in the light and electron microscope of identified Aplysia neurons and the processes after intrasomatic injection of e-[3H]fucose, Brain Research, 112 (1976) 251-281. 39 Weber, K. and Osborn, M., The reliability of molecular weight determinations by dodecyl sulfatepolyacrylamide gel electrophoresis, J. biol. Chem., 244 (1969) 4406-4412. 40 Williams, M. A., The assessment of electron microscopic autoradiographs, Advanc. opt. electr. Micros., 3 (1969) 219-272.