Journal of Neurocytology 21, 707-716 (1992)

G2-Acetylcholinesterase is presynaptically localized in Torpedoelectric organ J. E I C H L E R 1,2, I. S I L M A N 2 a n d L. A N G L I S T E R 1.

1 Department of Anatomy and Embryology, Hebrew University-Hadassah Medical School, Jerusalem 91010, Israel 2 Department of Neurobiology, The Weizmann Institute of Science, Rehovot 76100, Israel Received 30 January 1992; revised 1 June 1992; accepted 15 June 1992

Summary In Torpedo electric organ, much of the acetylcholinesterase (ACHE)is a globular dimer (G2), anchored to the plasma membrane via covalently attached phosphatidylinositol and selectively solubilized by a bacterial phosphatidylinositol-specific phospholipase C. While the structure of this form of the enzyme is well-established, the ultrastructural localization of G2-AChE is still unclear. Selective solubilization with phosphatidylinositol-specific phospholipase C was, therefore, combined with immunocytochemistry at the electron microscope level, in order to localize G2-AChE in electric organ of Torpedo ocellata. Thin sections of electric organ were labelled with antibodies raised against Torpedo ACHE, followed by gold-conjugated second antibodies, before or after exposure to the phospholipase. For comparison, the location of AChE was examined using histochemical methods. We show that (1) immunolabelling is concentrated in the synaptic clefts between nerve terminals and the innervated face of the electrocyte; (2) this labelling co-localizes with AChE histochemical reaction products; and (3) prior exposure to the phospholipase causes a decrease in AChE-associated labelling. Quantitative analysis of immunolabelling in the synaptic clefts shows that the phospholipase treatment had reduced primary labelling at or adjacent to the presynaptic membrane. Together with our earlier biochemical and immunofluorescent evidence, these results support our previous assignment of a neuronal and synaptic localization for GR-AChEin Torpedo electric organ.

Introduction Acetylcholinesterase (ACHE, EC 3.1.1.7) occurs in various excitable tissues of the elasmobranch, Torpedo, as a globular hydrophobic subunit dimer, G2 (Witzemann & Boustead, 1981; Futerman et al., 1987). This dimer was identified as one of the first members of a by now extensive list of proteins which are anchored to the plasma membrane by the diglyceride moiety of a single phosphatidylinositol (PI) residue which is covalently attached via an oligoglycan extension to the carboxyl-terminus of the protein (Low et aI., 1986; Silman & Futerman, 1987). Such proteins can often be solubilized by a PI-specific phospholipase C (PIPLC) of bacterial origin (Low et al., 1986). We have previously shown that this is the case for G2-AChE in a number of excitable tissues of Torpedo, including the electric organ, electromotor nerve and back muscle. Thus, the Ga dimer was selectively solubilized by PIPLC in biochemical studies using tissue homogenates (Futerman et al., 1987); these results were confirmed by immunocytochemical studies involving cryostat sections at the LM level (Eichler et al., 1990). It should be noted that the sensitivity to PIPLC is an * To whom correspondenceshouldbe addressed 03004864/92 $03.00 +.12 9 1992 Chapman and Hall Ltd

acquired property, since both the biochemical and immunocytochemical approaches show that the corresponding AChE dimer in the electric lobe is almost completely resistant to PIPLC (Futerman et al., 1987; Eichler et al., 1990). Unexpectedly, subcellular fractionation studies showed that a substantial portion of Torpedo electric organ AChE copurifies with the synaptosomal fraction; the AChE in this fraction is an externally-oriented hydrophobic dimer (Witzemann & Boustead, 1981; Morel & Dreyfus, 1982; Li & Bon, 1983) and is sensitive to PIPLC (Futerman et aI., 1985b). These data suggest that a large part of the PI-anchored dimer is attached to the presynaptic plasma membrane. Immunocytochemical studies using antibody preparations which recognize G2-AChE have reported results consistent with the concep~ of a presynaptic localization for this form of the enzyme (Kushner et aI., 1987; Abramson et aL, 1989; Mailly et al., 1989). These findings could not, however, conclusively establish the localization of G2-AChE, as the epitopes recognized by the various anti-AChE antibodies are expressed on molecules

708 o t h e r t h a n G2-AChE. F u r t h e r m o r e , these studies w e r e carried out m o s t l y at the LM level; as such, t h e y could not p r o v i d e the h i g h d e g r e e of resolution n e c e s s a r y for precise localization. W e h a v e recently d e v e l o p e d a n o v e l a p p r o a c h for the localization of G2-AChE in the Torpedo electrom o t o r s y s t e m (Eichler et aI., 1990). By c o m b i n i n g i m m u n o f l u o r e s c e n c e w i t h selective r e m o v a l of this f o r m of the e n z y m e b y the action of PIPLC, w e h a v e s h o w n a synaptic a n d n e u r o n a l localization for G2ACHE. To define m o r e precisely the localization of GR-AChE, w e n o w c o m b i n e the resolution offered b y the EM w i t h selective r e m o v a l of this f o r m of the e n z y m e b y PIPLC in electric o r g a n tissue of Torpedo ocellata, a n d , s h o w that the b u l k of G2-AChE is localized adjacent to the p r e s y n a p t i c m e m b r a n e .

Materials and methods MATERIALS Live specimens of Torpedo ocellata, caught off the shore of Tel Aviv, were generously donated by Dr R. Rahamimoff (Hebrew University-Hadassah Medical School, Jerusalem) and Dr D. Michaelson (University of Tel Aviv, Israel). Freshly-dissected electric organs were cut into small cubes and further processed for electron microscopy (see below). Torpedo Ringer's solution was prepared according to Sealock and Kavookjian (1980), except that 200-300 mM sucrose was included so as to maintain osmolarity at --1000mosM. Frozen Torpedo californica electric organ tissue (Marinus, Long Beach, CA) was stored at - 7 0 ~C. Monoclonal antibody (mAb) 4E7 was a gift from Dr B. P. Doctor (Walter Reed Army Institute of Research, Washington, DC), and a polyclonal anti-AChE preparation was donated by Dr Palmer Taylor (University of California, San Diego, La Jolla, CA). The characteristics of both these antibody preparations have been described previously (Doctor et al., 1983; Eichler et aI., 1990). A polyclonal antibody preparation, raised against Torpedo AChR was provided by Dr Sara Fuchs (Weizmann Institute of Science, Rehovot). PIPLC, purified from Staphylococcus aureus or from Bacillus thuringiensis (Low et al., 1988), was a gift from Dr M. G. Low (College of Physicians and Surgeons, Columbia University, New York, NY). PIPLCsolubilized and high-salt-soluble T. californica AChE were purified as described previously (Futerman et al., 1985a). SDS-PAGE reagents and molecular weight markers were obtained from Bio-Rad Laboratories (Cambridge, MA). Nitrocellulose paper (BA 85 membrane filters, 0.45 ~m) was from Schleicher & Schuell (Dassel, Germany). Horseradish peroxidase-conjugated goat anti-rabbit or goat anti-mouse IgG were from Jackson I.mmunoresearch (West Grove, PA). Colloidal gold (5 nm) conjugated to goat anti-rabbit or goat anti-mouse IgG was obtained from Janssen Pharmaceutica (Beerse, Belgium). 3,3'-diaminobenzidine (DAB) and bovine serum albumin (BSA) were from Sigma (St. Louis, MO), as was acetylthiocholine iodide. [3H]ACh was purchased from Amersham International (Amersham, UK), while [3H]diisopropylfluorophosphate (DFP) was obtained from New England Nuclear (Boston, MA).

EICHLER, SILMAN and ANGLISTER METHODS

Protein extraction Frozen T. ocellata electric organ tissue (50 g) was homogenized in 2.5 volumes of 2 MMgC12, 1 MNaC1, 1% Triton X-100, 1 mM EDTA, 10 mM Tris, pH 7.0 in an Omnimixer homogenizer (OCI Instruments, Waterbury, CT). The homogenate was centrifuged for 90 min at 100 000 g in a Beckman L7-55 ultracentrifuge using a Ti-35 rotor. The supernatant was extensively dialyzed against 100mM NaC1, 0.1% Triton X-100, 10 mM Tris, pH 8.0. A 1 : 10 dilution of the supernatant was used for SDS-PAGE studies (see below). Incubation of electric organ homogenates with PIPLC was performed as described previously (Futerman et al., 1985a).

SDS-PAGE and Western blots Torpedo electric organ extracts and purified asymmetric (A) or globular (G) AChE preparations were analyzed by SDSPAGE on 5-15% gels according to Futerman and colleagues (1985a). Proteins were transferred to nitrocellulose paper (Towbin & Gordon, 1984), quenched in blocking buffer (PBS, 0.5% Tween-20, 0.04% NaN3) containing 10% low fat milk, and exposed to the appropriate antibody preparation, at a 1:500 to 1 : 1000 dilution, for2 h at room temperature. In some cases, the nitrocellulose paper was incubated for 15 min in the fixative solutions used in the electron microscopy studies (see below), before being quenched and probed with antibodies. Antigen-antibody complexes were then detected using horseradish peroxidase-conjugated goat antirabbit or anti-mouse antibodies and the product of the DAB/H202 reaction.

[3H]DFP-Iabelling of AChE catalytic sites In order to label AChE catalytic sites, 100 ~1 aliquots of Torpedo electric organ extract (crude protein extract or PIPLC-solubilized fraction) were incubated with 4 I~Ci of [3H]DFP for I h at room temperature, followed by SDSPAGE. Autofluorography was performed as previously described (Laskey & Mills, 1975), except that 0.4% PPO in acetic acid/xylene/ethanol (1.8:1:2, v/v/v) replaced 20% PPO in DMSO.

Tissue preparation for electron microscopy Fresh T. ocellata electric organ tissue was dissected out, trimmed to approximately 2 x 3 m m cubes and weighed. The tissue samples were incubated in 0.5 ml of Ringer's solution in the presence or absence of 10-20 U of PIPLC for 4.5 h at room temperature. The bathing solutions were collected and assayed for AChE content (see below). Tissue pieces were fixed in Ringer's containing 4% glutaraldehyde (2 h, 4~ C) or in periodate-lysine-paraformaldehyde fixative (PLP) (McLean & Nakane, 1974) (6 h, 4~C) and incubated in 1% NH4C1 in 0.1 M phosphate (pH 7.2) for 3 x 15 min, to quench reactive aldehyde groups (Brown & Farquhar, 1984). Next, the samples were rinsed with 0.1 M cacodylate (pH 7.2), dehydrated with ethanol and finally embedded in LR gold acrylic resin (Ted Pella Inc., Irvine, CA) and polymerized under ultraviolet light.

Electron microscopic immunocytochemistry Thin sections of embedded tissue samples, some of which had been previously exposed to PIPLC, were cut and

Presynaptic localization of

Torpedo G2-acetylcholinesterase

mounted on nickel grids. These sections were incubated for 20 min at room temperature in buffer A (PBS, 1% Tween-20, 0.5% BSA, 1% gelatin, 1% glycine), in order to reduce non-specific labelling. Next, the sections were incubated for 2 h at room temperature with the various antibodies at the following dilutions, prepared in buffer A:1:150 for antiAChE serum, 1 : 25 for polyclonal anti-AChR antibodies, and 1:150 for mAb 4E7. In all experiments, normal rabbit or mouse sera, at the corresponding dilutions, were used as negative controls. Afterwards, grids were washed for 6 • 5 rain in PBS, and probed with 5 nm gold-conjugated goat anti-rabbit or anti-mouse IgG diluted 1 : 25 in buffer A, for i h at room temperature. After 3 x 5 min PBS washes the grids were rinsed in distilled water and counterstained in 2% uranyl acetate for 20 rain at room temperature. Finally, the grids were extensively washed in water, dried and examined at 80 kV in a Phillips 410 EM.

Statistical analysis of antibody labelling Photographs of immunolabelled tissue sections were prepared and the density of gold particles at the presynaptic membrane was determined as follows: the synaptic cleft was divided transversely (i.e. parallel to the presynaptic membrane) into 1 mm segments (corresponding to 4.6 nm at this magnification). The number of gold particles per segment was counted and then normalized per unit length of presynaptic membrane to obtain densities. The average density of particles in each segment, representing a given distance from the presynaptic membrane, was then compared in control and PIPLC-treated samples. Photographs from a total of 48 samples, each containing an average length of presynaptic membrane of 151 _+9 mm, were measured. Analysis of the dispersed labelling observed at the nonsynaptic face of the nerve terminals was determined by counting the average number of gold particles in control and PIPLC-treated samples in the area enclosing the nerve terminal membrane and its accompanying Schwann cell.

Electron microscopic histochemistry Fresh T. ocellataelectric organ tissue was excised in Ringer's solution and individual electroplaque columns were dissected out. The pieces of tissue were fixed in 1% glutaraldehyde in Ringer's for 25 min, at room temperature, prior to histochemical staining for AChE activity according to Karnovsky and Roots (1964). The tissue samples were incubated in the presence of 0.2 mM substrate (acetylthiocholine iodide) for 30 min and, finally, in the presence of 2.0 mM substrate for I h. They were then rinsed in Ringer's solution, fixed for l h in 1% OsO4/0.1M phosphate (pH 7.2), dehydrated in ethanol, incubated in propylene oxide, embedded in a mixture of Epon-propylene oxide (1:1) and finally mounted flat in Epon (Anglister, 1991). Ultrathin sections were grid-stained routinely with uranyl acetate and lead citrate. To enhance the exposure of the tissue to the reagents, samples were, in some cases, transferred after the aldehyde fixation into PBS containing 2.3 M sucrose (overnight, 4~ C). These samples were then frozen in isopentane/ liquid N2 and 30 ~m cryosections (cut on a Frigocut 2800E cryostat, Reichertqung, Nussloch, Germany) were stained for AChE activity for 10 rain in Karnovsky's mixture containing 0.35 mM acetylthiocholine iodide. Further processing for EM was as described above.

709

Determination of AChE activity AChE activity was determined according to Johnson and Russell (1975), using [3H]ACh as substrate. Results

Characterization of antibodies In Torpedo electric organ, AChE exists both as an asymmetric (A) molecular species, containing 66 kDa catalytic subunits, or as a globular (G) form, containing 63 kDa subunits (Futerman et aI., 1985a). To test the abilities of the anti-AChE polyclonal antibodies and mAb 4E7 to recognize purified AChE from Torpedo electric organ, both affinity-purified A- and PIPLCsolubilized G-AChE subunits were e x a m i n e d b y Western blotting after SDS-PAGE, p e r f o r m e d u n d e r reducing conditions (Fig. 1). The polyclonal preparation recognized both AChE molecular forms equally well, while mAb 4E7 recognized the G form preferentially (Fig. 1A, B,C), as previously r e p o r t e d by Doctor and colleagues (1983). The abilities of both the anti-AChE polyclonal and m o n o c l o n a l antibodies to react with a crude protein extract of Torpedo electric organ tissue were next e x a m i n e d b y SDS-PAGE u n d e r reducing conditions followed b y Western blotting. Both the anti-AChE polyclonal a n t i s e r u m a n d mAb 4E7 recognized a protein b a n d at 63 kDa, as well as a 135 kDa b a n d (Fig. 1D,E). [3H]DFP-labelling of AChE catalytic sites s h o w e d that these b a n d s c o r r e s p o n d to e n z y m e m o n o m e r s and dimers, respectively (Fig. 1H). An a n t i s e r u m raised against the AChR labelled, as expected, b a n d s of 40, 45, 60, and 69 kDa (Bartfeld & Fuchs, 1979), c o r r e s p o n d i n g to the AChR o~-, [3-, ~- and 8- subunits, respectively (not shown). Proteins released from Torpedo electric organ homogenates s u b s e q u e n t to incubation with PIPLC were also e x a m i n e d by Western blotting. The polyclonal antibodies a n d mAb 4E7 reacted strongly with a 63 kDa b a n d (Fig. 1F, G). Binding to a 135 kDa b a n d was also observed. As above, DFP-labelling studies confirmed that the 63 and the 135kDa b a n d s corr e s p o n d to AChE m o n o m e r s and dimers, respectively (not shown). Thus, in Torpedo electric organ, AChE is the main protein recognized by the anti-AChE antibodies used. In crude protein extracts of Torpedo electric organ, a faint doublet b a n d (> 200 kDa) was also recognized by both antibodies. This b a n d was not, h o w e v e r , released b y PIPLC, n o r was it labelled by [3H]DFP (compare Fig. 1D,E with F,G and H). In a separate set of experiments, using T. californica as the source of electric organ tissue, similar results were obtained; the polyclonal antibodies a n d mAb 4E7 recognized the same protein b a n d s as in T. ocellata electric organ, in both the crude protein extracts and in

710

EICHLER, SILMAN and ANGLISTER the PIPLC-released supernatant (Fig. 1A-G). Finally, [3H]DFP labelled the same protein bands in both species (not shown).

Histochemical detection of AChE By the Karnovsky and Roots m e t h o d (1964), staining for AChE could be detected both within t h e postsynaptic folds and in the primary synaptic cleft (Fig. 2). Precipitate was also observed on the non-synaptic side of the nerve terminals, but the a m o u n t of precipitate was m u c h lower than observed in the synaptic cleft or folds (Fig. 2A,B). In the synaptic cleft, reaction product appeared to be concentrated at the presynaptic face but also stained the postsynaptic face (Fig. 2D). Very little staining of extra-synaptic regions (Fig. 2A) or of the non-innervated face of the electrocyte could be detected (Fig. 2C).

Efficiency of PIPLC-mediated release of AChE

Fig. 1. Binding of anti-AChE antibodies to purified Torpedo AChE and to electric organ extracts. Samples of purified AChE and tissue extracts, prepared as described in Methods, were separated using 5-15% SDS-PAGE. Western blots were then probed with the antibody preparations. Samples included: purified A-AChE (A), G2-AChE (B,C), electric organ extracts (D,E,H) and supernatants of electric organ homogenates spun (100 000 g, 90 min) after incubation with PIPLC (0.5 U g 1, 16 h, RT) (F,G). Blots were reacted with either the polyclonal preparation (A,B,D,F) or mAb 4E7 (C,E,G). The polyclonal antibodies recognized catalytic subunit polypeptides of purified A-AChE (A) and PIPLC-solubilized G2-AChE (B), while mAb 4E7 recognized only the latter (C). In electric organ extracts, the polyclonal preparation (D) and mAb 4E7 (E) both recognized protein bands of 63 and 135 kDa, as well as a > 200 kDa doublet. In the supernatants of PIPLC-treated tissue homogenates, only protein bands of 63 and 135 kDa were labelled by the polyclonal antibodies (F) and mAb 4E7 (G). (H) Autoradiogram of Torpedo electric organ extract, pretreated with [3H]DFP. The same protein bands (63 and 135 kDa) were radiolabelled, corresponding to enzyme monomers and dimers, respectively. Arrowheads indicate the migration of standard molecular weight markers on the right, and the catalytic subunits of A- and G2-AChE on the left.

The extent of PIPLC-induced AChE release was determ i n e d by comparing the levels of AChE activity in the bathing media of small tissue blocks incubated in the presence or absence of PIPLC, prior to fixation and subsequent processing for EM. These activities were then normalized per unit weight (grams of wet tissue). Solutions from samples to which PIPLC had been a d d e d contained almost 20 times more AChE activity t h a n did control solutions (17 + 2 SEM, n = 3). Under the conditions employed, the a m o u n t of AChE released by the phospholipase treatment represents only a small percentage of the PIPLC-sensitive enz y m e f o u n d in the electric organ (Futerman et al., 1985b). Solubilization by PIPLC was limited, both due to restricted penetration of the phospholipase into the tissue blocks and due to the fact that a short incubation time was a requirement for retention of the ultrastructural features of the tissue. Consequently, only the outermost tissue slices were u s e d in immunocytochemical studies.

Immunocytochemical labelling of Torpedo electric organ by anti-AChE polyclonal antibodies Using an experimental approach successfully employed to localize G2-AChE at neuronal and synaptic sites in Torpedo excitable tissues at the LM level (Eichler et aI., 1990), we attempted to localize this form of the

Fig. 2. Histochemical localization of AChE activity in Torpedo ocellata electric organ. Cryosections of electric organ were fixed and stained by the Karnovsky procedure (details in Methods). (A,B) Innervated surface of the electrocyte. Staining for AChE appeared on the innervated side of the electrocyte, where stain was concentrated in the synaptic cleft (arrows) and in the postsynaptic folds (arrowheads). Reaction products also appeared on the non-synaptic face of the nerve terminals and around their capping Schwann cells (open arrows). Arrows and arrowheads are positioned inside the electrocyte, the open arrows are in the extracellular space. Note that only little staining is observed on the extra-synaptic surfaces of the electrocyte (*). (C) The opposite, non-innervated side of the electrocyte. Very faint staining is associated with this surface. (D) Magnified view of part of the synaptic cleft shown in (A), illustrating that staining of the presynaptic surface was stronger than of the postsynaptic surface. Scale bars: (A,C) 0.5 p~m;(B) 0.3 ~m; (D) 0.2 ~m.

Presynaptic localization of TorpedoG2-acetylcholinesterase

711

712

EICHLER, SILMAN and ANGLISTER A

o

nerve >
< electrocyte

I"-ICi []

+ PIPLC

9 9

o

0.5 0.0. -2 -1 0 1 2 3 4 5 6 7 8 9 distance from presynaptic membrane (mm)

B

O

O

Fig. 3. Effect of PIPLC on the immunogold labelling of AChE in the synaptic cleft by polyclonal antibodies. Electron micrographs of thin sections of embedded tissue samples (800 A) labelled with a 1:150 dilution of polyclonal antiAChE serum and probed with 5 nm gold-conjugated goat anti-rabbit IgG. (A) Untreated electric organ tissue, control. Immunolabelling could be detected at the presynaptic membrane (arrows), with a lesser degree of labelling noted at the postsynaptic face (open arrows). Note that the nerve terminal is detached from the postsynaptic membrane, thus emphasizing the preferential labelling of the presynaptic face. (B) Tissue after exposure to PIPLC (details in Methods). Incubation with PIPLC prior to labelling greatly reduced the number of gold particles in the synaptic cleft, specifically at the presynaptic membrane (arrows). Most of the remaining gold particles observed in the synaptic cleft were found close to the postsynaptic membrane (open arrows). Scale bar: 0.1 btm.

e n z y m e at the EM level. In the p r o c e d u r e d e v e l o p e d , PIPLC t r e a t m e n t h a d to p r e c e d e a n y fixation. Moreover, o s m i u m fixation w a s o m i t t e d altogether, as such t r e a t m e n t h i n d e r s b o t h a n t i b o d y b i n d i n g a n d resin p o l y m e r i z a t i o n (Stirling, 1990). As a result, certain structures, such as vesicles w i t h i n the n e r v e terminals, w e r e not w e l l - p r e s e r v e d , a l t h o u g h b o t h the n e r v e terminal a n d electric o r g a n p l a s m a m e m b r a n e s (as well as the basal l a m i n a lying b e t w e e n t h e m ) w e r e well m a i n t a i n e d . W e verified that fixation did not significantly affect a n t i b o d y binding: the antibodies labelled A C h E s u b u n i t s of Torpedo electric o r g a n extracts, s e p a r a t e d as in Fig. 1, b u t s u b s e q u e n t l y fixed after

9

,.ro

synaptic cleft >< electrocyte nerve >< 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1-__ 0.0 2 -2 -1 0 1 2 3 4 5 6 7 8 9 distance from presynaptic membrane (mm)

i/

Fig. 4. Statistical analysis of the effect of PIPLC on the density of labelling at the presynaptic membrane. Stretches of synaptic cleft regions in sections of uniform thickness (800 A) were photographed and divided transversely into 1ram segments (corresponding to 4.6nm, details in Methods). The number of gold particles per segment was counted and normalized per unit length of presynaptic membrane. Densities are expressed as number of particles per 0.4 txm of membrane (error bar, S~M). The density of particles at a given distance from the presynaptic membrane (a segment) was then compared in control (n = 36) and PIPLC-treated samples (n = 12). Samples were obtained from electric organs of two animals in three labelling experiments. (A) Statistical analysis of the labelling patterns of the anti-AChE polyclonal antibodies showed that two to three times more gold particles were counted at the presynaptic membrane (segment: 0 ram) than at the postsynaptic membrane (segment: 6ram) ( p < 0.01). However, upon treatment with PIPLC, the binding pattern was greatly altered, the number of particles at the presynaptic membrane being reduced two to three-fold (+p < 0.01, *p < 0.05). Labelling within the deft was not significantly reduced nor was labelling at the postsynaptic membrane. (B) Statistical analysis of the labelling patterns of the anti-AChR antibodies showed that labelling of the acetylcholine receptor was concentrated at the post-synaptic membrane (n = 14).

Presynaptic localization of Torpedo G2-acetylcholinesterase transfer to nitrocellulose according to the same fixation regime used in the immunocytochemical procedure. In order to examine the fine distribution of AChE in Torpedo electric organ synapses, the tissue was exposed to polyclonal anti-Torpedo AChE antibodies (Fig. 3). Substantial immunogold labelling could be detected at the presynaptic face (Fig. 3A), with a lesser degree of labelling being noted in the synaptic cleft and at the base of the postsynaptic folds (not shown). The labelling pattern of electric organ by the polyclonal antibodies was greatly affected by incubation with PIPLC prior to the fixation and subsequent labelling steps. In sections so treated, the number of gold particles in the synaptic cleft was greatly reduced in comparison with sections not exposed to the phospholipase, particularly at the presynaptic face (compare Fig. 3A and B). Most of the remaining gold particles observed in the synaptic cleft were found close to the postsynaptic membrane, on the basal lamina, or within the postsynaptic invaginations. Figure 4A displays quantitative analyses of the labelling patterns obtained with the anti-AChE antibodies and of the effect of PIPLC on these patterns. In the case of tissue labelled with the polyclonal preparation, two to three times more gold particles were counted at the presynaptic membrane than at the postsynaptic membrane (Fig. 4A, segments: - 2 to i mm). Treatment with PIPLC significantly reduced the number of particles at the presynaptic membrane. Labelling within the synaptic cleft was not significantly altered, while labelling at the postsynaptic membrane was even slightly increased. It could be calculated that the PIPLC treatment removed almost 60% of the epitopes recognized by the polyclonal antiserum at the presynaptic face, a value in excellent agreement with previous reports concerning the extent of AChE release by PIPLC (Futerman et aI., 1985b; Eichler et al., 1990). To confirm that our approach indeed provides sufficient resolution to distinguish between pre- and postsynaptic localization, we examined immunogold labelling with antiserum directed against the AChR, a molecule known to be exclusively localized at the poststynaptic membrane (Matthews-Bellinger & Salpeter, 1978; Sealock et al., 1984). As expected, examination of the labelling pattern showed that gold particles were almost completely restricted to the synaptic clefts, and little labelling of the nerve terminals or of the non-innervated face of the electroplaque was observed. Quantitative analysis revealed that the majority of gold particles lay within 5 n m on either side of the postsynaptic membrane (Fig. 4B). Approaching the presynaptic membrane, a distinct drop in the number of particles was recorded. Given that the length of the antibodies themselves may slightly distance the gold particle from the AChR, these results

713

reflect the exclusive postsynaptic localization of the receptor (Sealock et al., 1984). Dispersed labelling with polyclonal anti-AChE antibodies could be seen on the non-synaptic face of the nerve terminals, which at times were seen in association with Schwann cell processes (not shown). This labelling did not seem to be associated with any particular cellular structure, but rather with the extracellular matrix encasing the nerve terminals and Schwann cells. In contrast to the situation found in the synaptic cleft, PIPLC reduced this labelling by only 30%. No labelling of the non-innervated face of the electroplaque was observed.

Immunocytochemical labelling of Torpedo electric organ by mAb 4E7 In spite of using concentrations of mAb 4E7 severalfold higher than required for successful labelling with the polyclonal antibodies, or for its successful use in immunofluorescence (Eichler et al., 1990), no labelling of the synaptic cleft was achieved by mAb 4E7, regardless of whether the tissue had been fixed in glutaraldehyde or in the milder PLP fixative (not shown). Gold particles were diffusely spread across the non-synaptic side of the nerve terminal as well as being found at the non-innervated face of the electrocyte; this labelling was not affected by prior exposure to PIPLC. The low extent of synaptic cleft labelling by this antibody precluded quantitative analysis of labelling or of the effect of PIPLC. Discussion

AChE exists in a variety of molecular forms, differing in their hydrodynamic behaviour, degree of oligomerization, glycosylation patterns, and modes of anchoring (Silman & Futerman, 1987; Massouli6 & Toutant, 1988). The pattern of these forms may vary from tissue to tissue even in the same animalt (Silman et al., 1978). No functional explanation has yet been provided for these structural differences, since all the various molecular forms display similar catalytic activity (Vigny et al., 1978). It has been suggested, however, that their existence reflects the need of a synapse to localize AChE in such a way as to optimize termination of cholinergic transmission. Electrophysiological evidence has been provided to support this possibility (Federov, 1981). Direct experimental support for such a claim, based on immunocytochemical localization of the various molecular forms within the synapse, has been hampered by lack of formspecific antibodies (Brimijoin & Rakonczay, 1986). This limitation may be partially overcome by taking advantage of differential susceptibilities of the anchoring domains of the various molecular forms. G2-AChE has been shown to be sensitive to the actions of PIPLC in Torpedo electric organ (Futerman et al.,

714 1983). We have taken advantage of this phenomenon, combined with immunocytochemistry at the EM level, so as precisely to localize G2-AChE in Torpedo electric organ. Although histochemical staining revealed the bulk of the AChE to be concentrated within the synaptic cleft, lower levels could be detected between the nerve ending and the Schwann cell capping it, as well as around the Schwann cell itself. Such staining might reflect the presence of active enzyme, though we cannot rule out the possibility that it may also result from the diffusion of reaction product (see below). However, it should be noted that no staining developed at extra-synaptic sites, i.e. at sites devoid of nerve terminals or Schwann cells. It should be noted that this finding is in contrast to the considerable staining of extra-synaptic sites observed in Electrophorus electricus electric organ, where long stretches of the innervated face are not contacted by incoming nerve terminals (Bloom & Barrnett, 1966). Furthermore, no enzymatic activity was noted on the non-innervated face of the electrocytes. While useful for detecting active ACHE, such staining is inadequate for precise localization of the enzyme within the synaptic cleft, principally due to the well-documented phenomenon of precipitate diffusion (Tsuji, 1974). Moreover, this approach does not permit differentiation between the various molecular forms. It does, however, allow us to conclude that the bulk of electric organ AChE is found within the synaptic cleft. It also appeared that more staining occurred at the presynaptic side of the synaptic cleft as compared to the postsynaptic face. Immunocytochemistry using gold-labelled conjugates permitted more precise localization of specific elements within the synaptic cleft. Using this approach, we observed that labelling of AChR was confined to the postsynaptic membrane, confirming earlier results (Matthews-Bellinger & Salpeter, 1978; Sealock et al., 1984). Two different antibodies were used in the localization of ACHE: a polyclonal preparation and a monoclonal antibody, mAb 4E7 (Eichler et al., 1990). Both these antibody preparations recognize purified G2-AChE, while the polyclonals also recognize A-AChE (Doctor et al., 1983). When used to label protein extracts of Torpedo electric organ, both antibodies were able to recognize AChE dimers and monomers, as identified by DFP binding. Prior fixation of the tissue did not change these labelling patterns. Tissue sections labelled with the polyclonal antiAChE preparation displayed substantial labelling within the synaptic cleft. In addition, a dispersion of gold particles was noted over the area occupied by the non-synaptic face of the nerve terminal and the adjacent Schwann cell. In agreement with previous LM studies (Eichler et al., 1990), no labelling of the non-innervated face of the electrocyte was detected.

EICHLER, SILMAN and ANGLISTER Statistical analysis showed that up to three times more labelling could be detected at the presynaptic membrane, as compared with the rest of the synapse. As G2-AChE accounts for up to 70% of the AChE in Torpedo electric organ (Futerman et al., 1985b), this lends support to a predominantly presynaptic localization for this form of ACHE. When the effect of PIPLC on this labelling was examined, it was observed that the phospholipase substantially reduced binding at the presynaptic face, but not inside the synaptic cleft or at the postsynaptic face. Statistical analysis showed that in the vicinity of the presynaptic membrane, labelling by the polyclonal antibody was significantly reduced by PIPLC treatment, the decrease ranging between 2- and 3-fold. No significant decrease in labelling could be detected at the postsynaptic face of PIPLC-treated tissue. In fact, a slight increase could be observed, possibly due to the marked reduction in presynaptic G2-AChE, permitting more substantial labelling of A-AChE, believed to be localized at the basal lamina (for review, Anglister & McMahan, 1984; Silman & Futerman, 1987). Labelling of the nonsynaptic area was much less affected by PIPLC treatment. In tissue immunocytochemically labelled with mAb 4E7 and examined by EM, gold particles were detected throughout the area encompassing the non-synaptic face of the nerve terminal and, in many cases, also on Schwann cells encasing nerve terminals. Gold particles were seen at the non-innervated face of the electroplaque, from which AChE is known to be totally absent, in agreement with previous immunofluorescence studies (Abramson et al., 1989; Eichler et al., 1990). Practically no labelling could be detected in the synaptic cleft, as was reported by Abramson and colleagues (1989), leading these authors to localize the dimer to the non-synaptic surface of the nerve terminals and possibly to the Schwann cell processes. Although mAb 4E7 binds to G2-AChE subunits, as demonstrated in the Western blotting experiments (Fig. 1), as well as to the native dimer, as demonstrated by dimer-antibody complex formation in sucrose gradient centrifugation experiments (not shown), the observed immunogold labelling of tissue processed for EM cannot reflect binding to G2-AChE. This form of the enzyme was shown to be solubilized quantitatively by PIPLC (Futerman et al., 1983). Yet, this labelling was unaffected by prior exposure to the phospholipase. Moreover, as shown in the Western blotting experiments, G2-AChE is not the only protein in Torpedo electric organ recognized by mAb 4E7. Finally, histochemical localization of AChE showed that the vast majority of enzyme is found within the synaptic cleft. As G2-AChE accounts for the major part of AChE in Torpedo electric organ, one would expect this form to contribute at least some of the staining in the synaptic cleft. However, no such labelling was

715

P r e s y n a p t i c localization of Torpedo G2-acetylcholinesterase d e t e c t e d u s i n g m A b 4E7. W e c o n c l u d e t h a t m A b 4E7 c a n n o t be u s e d in t h e s e i m m u n o c y t o c h e m i c a l studies. T h e f u n c t i o n a l significance of the p r e s y n a p t i c locali z a t i o n of G 2 - A C h E in Torpedo electric o r g a n r e m a i n s to be clarified. I n Torpedo b a c k m u s c l e , the G2 f o r m a c c o u n t s for the b u l k of the A C h E activity ( W i t z e m a n n & B o u s t e a d , 1981) a n d it is largely P I P L C - s e n s i t i v e ( F u t e r m a n et al., 1987; Eichler et aI., 1990). In t h e s l o w tonic a n t e r i o r latissimus dorsi m u s c l e of the c h i c k e n , a G2 f o r m also p r e d o m i n a t e s , b o t h in the w h o l e m u s c l e (Silman et al., 1978) a n d w i t h i n t h e e n d p l a t e r e g i o n ( J e d r z e j c z y k et al., 1984). It will be of interest to establish w h e t h e r in t h e s e m u s c l e e n d p l a t e s too, t h e G2 f o r m of A C h E is localized o n t h e n e u r o n a l side of the s y n a p t i c cleft.

Acknowledgements T h e a u t h o r s w i s h to t h a n k Mrs Rachel C o h e n for h e r excellent technical assistance. This r e s e a r c h w a s s u p p o r t e d b y g r a n t s f r o m the Israel N a t i o n a l A c a d e m y of Sciences (to L.A. a n d I.S.), f r o m the U.S.-Israel Binational Science F o u n d a t i o n a n d f r o m the B r u n o G o l d b e r g E n d o w m e n t F u n d (to L.A.), f r o m the Association F r a n q a i s e - I s r a e l i e n n e de la R e c h e r c h e Scientifique et T e c h n o l o g i q u e (AFIRST) a n d f r o m t h e M i n e r v a F o u n d a t i o n (to I.S.). I.S. is B e r n s t e i n - M a s o n P r o f e s s o r of N e u r o c h e m i s t r y .

References ABRAMSON, S. N., ELLISMAN, M. H., DEERINCK, T. J., MAULET, Y., GENTRY, M. K., DOCTOR, B. P. & TAYLOR, P. (1989) Differences in structure and distribution of the molecular forms of acetylcholinesterase. Journal of Cell Biology 108, 2301-11. ANGLISTER, L. (1991) Acetylcholinesterase from the motor nerve terminal accumulates on the synaptic basal'lamina of the myofiber. Journal of Cell Biology 115, 755-64. ANGLISTER, L. & MCMAHAN, U. J. (1984) Extracellular matrix components involved in neuromuscular transmission and regeneration. In Basement Membranes and Cell Movement (edited by PORTER, R. & WHELAN, R.) pp. 16178. London: Pitman. BARTFELD, D. & FUCHS, S. (1979) Fractionation of antibodies to acetylcholine receptor according to antigenic specificity. FEBS Letters 105, 303-6. BLOOM, F. E. & BARRNETT, R. J. (1966) Fine structural localization of acetylcholinesterase in electroplaque of the electric eel. Journal of Cell Biology 29, 475-95. BRIMIJOIN, S. & RAKONCZAY, Z. (1986) Immunology and molecular biology of the cholinesterases: current results and prospects. International Review of Neurobiology 28, 363-410. BROWN, W. J. & FARQUHAR, M. G. (1984) The mannose-6phosphate receptor for lysosomal enzymes is concentrated in cis golgi cisternae. Cell 36, 295-307. DOCTOR, B. P., CAMP, S., GENTRY, M. K., TAYLOR, S. S. & TAYLOR, P. (1983) Antigenic and structural differences in the catalytic subunits of the molecular forms of acetylcholinesterase. Proceedings of the National Academy of Science (USA) 80, 5767-71. EICHLER, J., SILMAN, I., GENTRY, M. K. & ANGLISTER, L. (1990) Immunocytochemical localization of phosphatidylinositol-anchored acetylcholinesterase in excitable tissues of Torpedo ocellata. Molecular Brain Research 8, 213-18. FEDEROV, V. V. (1981) Postsynaptic endplate ionic currents in fast and slow muscle fibres of chickens, effect of cholinesterase inhibition. Neurophysiology 13, 396-7. EUTERMAN, A. H., LOW, M. G. & SILMAN, I. (1983) A hydrophobic dimer of acetylcholinesterase from Torpedo californica electric organ is solubilized by phosphatidy-

linositol-specific phospholipase C. Neuroscience Letters 40, 85-9. FUTERMAN, A. H., FIORINI, R. M., ROTH, E., LOW, M. G. & SILMAN, I. (1985a) Physicochemical behaviour and structural characteristics of membrane-bound acetylcholinesterase from Torpedo electric organ. Biochemical Journal 226, 369-77. FUTERMAN, A. H., LOW, M. G., MICHAELSON, D. M. & SILMAN, I. (1985b) Solubilization of membrane-bound acetylcholinesterase by a phosphatidylinositol-specific phospholipase C. Journal of Neurochemistry 45, 1487-94. FUTERMAN, A. H., RAVIV, D., MICHAELSON, D. M. & SILMAN, I. (1987) Differential susceptibility to phosphatidylinositol-specific phospholipase C of acetylcholinesterase in excitable tissues of embryonic and adult Torpedo ocellata. Molecular Brain Research 2, 105-12. JEDRZEJCZYK, J., SILMAN, I., LAI, J. & BARNARD, E. A. (1984) Molecular forms of acetylcholinesterase in synaptic and extrasynaptic regions of avian tonic muscle. Neuroscience Letters 46, 283-9. JOHNSON, C. D. & RUSSELL, R. L. (1975) A rapid simple radiometric assay for cholinesterase suitable for multiple determinations. Analytical Biochemistry 64, 229-38. KARNOVSKY, M. J. & ROOTS, L. (1964) A "direct-coloring" thiocholine method for cholinesterases. Journal of Histochemistry and Cytochemistry 12, 219-21. KUSHNER, P. D., STEPHENSON, D. T., STERNBERG, H. & WEBER, R. (1987) Monoclonal antibody Tor 23 recognizes a determinant of a presynaptic acetylcholinesterase. Journal of Neurochemistry 48, 1942-53. LASKEY, R. A. & MILLS, A. D. (1975) Quantitative film detection of 3H and 14C in polyacrylamide gels by fluorography. European Journal of Biochemistry 56, 335-41. LI, Z. Y. & BON, C. (1983) Presence of a membrane-bound acetylcholinesterase form in a preparation of nerve endings from Torpedo marmorata electric organ. Journal of Neurochemistry 40, 338-49. LOW, M. G., FERGUSON, M. A. J., FUTERMAN, A. H. & SILMAN, I. (1986) Covalently attached phosphatidylinositol as a hydrophobic anchor for membrane proteins. Trends in Biochemical Sciences 11, 212-15.

716 LOWI M. G., STIERNBERG, J., WANECK, G. L., FLAVELL, R. A. & KINCADE, P. W. (1988) Cell-specific heterogeneity in sensitivity of phosphatidylinositol-anchored m e m b r a n e antigens to release by p h o s p h o l i p a s e C. Journal of hnmunological Methods 113, 101-11. MAILLY, P., YOUNI~S-CHENOUFI, A. B. & BON, S. (1989) The monoclonal antibodies Elec-39, HNK-1 and NC-1 recognize structures in the nervous system and muscles of vertebrates. Neurochemistry International 15, 517-30. MASSOULII~, J. & TOUTANT, J. P. (1988) Vertebrate cholinesterases: structure a n d types of interaction. In Handbook of Experimental Pharmacology, vol. 86 (edited by WHITTAKER, V. P.) pp. 167-224. Berlin: Springer-Verlag. MATTHEWS-BELLINGER, J. & SALPETER, M. M. (1978) Distribution of acetylcholine receptors at frog neuromuscular junctions with a discussion of some physiological implications. Journal of Physiology 279, 197-213. MCLEAN, I. W. & NAKANE, P. K. (1974) Periodate-lysinep a r a f o r m a l d e h y d e fixative. A n e w fixative for i m m u n o electron microscopy. Journal of Histochemistry and Cytochemistry 22, 1077-83. MOREL, N. & DREYFUS, P. (1982)Association of acetylcholinesterase with the external surface of the presynaptic plasma m e m b r a n e in Torpedo electric organ. Neurochemistry International 4, 283-8. SEALOCK, R. & KAVOOKJIAN, A. (1980) Postsynaptic distribution of acetylcholine receptors in electroplax of the Torpedine ray Narcine brasiliensis. Brain Research 190, 81-93. SEALOCK, R., WRAY, B. & FROEHNER, S. C. (1984) Ultra-

EICHLER, S I L M A N and A N G L I S T E R structural localization of the Mr 43,000 protein and the acetylcholine receptor in Torpedo postsynaptic membranes using monoclonal antibodies. Journal of Cell Biology 98, 2239-44. SILMAN, I., LYLES, J. M. & BARNARD, E. A. (1978) Intrinsic forms of acetylcholinesterase in skeletal muscle. FEBS Letters 94, 166-70. SILMAN, I. & FUTERMAN, A~ H. (1987) Modes of attachment of acetyIcholinesterase to the surface membrane. European Journal of Biochemistry 170, 11-22. STIRLING, J. W. (1990) I m m u n o - and affinity probes for electron microscopy: a review of labeling and preparation techniques. Journal of Histochemistry and Cytochemistry 38, 145-57. TOWBIN, H. & GORDON, J. (1984) Immunoblotting and dot i m m u n o b i n d i n g - c u r r e n t status and outlook. Journal of Immunological Methods 72, 313-40. TSUJI, S. (1974) O n the chemical basis of thiocholine m e t h o d s for demonstration of acetylcholinesterase activities. Histochemistry 42, 99-110. VIGNY, M., GISIGER, V. & MASSOULIt~, J. (1978) "Nonspecific" cholinesterase and acetylcholinesterase in rat tissues: molecular forms, structural a n d catalytic properties, a n d significance of the two e n z y m e systems. Proceedings of the National Academy of Science (USA) 75, 2588-92. WITZEMANN, V. & BOUSTEAD, C. (1981) Distribution of acetyJcholinesterase molecular forms in brain, nerve and muscle tissue of Torpedo marmorata. Neuroscience Letters 26, 313-18.