Journal of Neuroscience Research 30:259-265 (1991)

Identification of C ytoskeletal, Focal Adhesion, and Cell Adhesion Proteins in Growth Cone Particles Isolated From Developing Chick Brain C. Cypher and P.C. Letourneau Department of Cell Biology and Neuroanatomy , University of Minnesota, Minneapolis

Growth cones are intimately involved in determining the direction and extent of neurite elongation during development. They are able to monitor their environ- ment and respond to it by undergoing directed mo- tility. We have isolated a fraction enriched in growth cone particles from embryonic chick brain. Assayed by immunoblots, this fraction is enriched in GAP-43, and contains the cytoskeletal proteins actin, myosin 11, neurofilament protein, tubulin, kinesin, and dy- namin. All of the major components of focal adhe- sions are also present: alpha-actinin, vinculin, talin, and integrin. In addition to integrin, we also identify the cell adhesion molecules A-CAM, L1, fibronectin, and laminin in these particles. This preparation of isolated growth cone particles may be a useful model system for studying growth cone adhesion and motil- ity.

Key words: growth cones, cytoskeleton, cell adhe- sion

INTRODUCTION The leading tips of neurites, the growth cones, are

versatile and dynamic structures. They are specialized to perform two functions essential to neurite growth and neuronal regeneration. First, they perceive their environ- ment and detect changes in it. Second, they respond appropriately to their environment and use that informa- tion to undergo directed motility. While performing these sensory and motile functions, growth cones trans- duce extracellular signals into changes in cytoskeletal organization and function,

Many studies have examined the behavior of growth cones and the distributions of known cytoskeletal components within them. More recently, attention has turned to the proteins that interact directly with the ex- tracellular substrate; the cell adhesion molecules, and their receptors on and within growth cone membranes. This study reports the first demonstration that proteins

found in focal adhesions, and extracellular cell adhesion proteins, are present in isolated growth cone particles (GCPs).

As they move, growth cones extend filopodia and lamellipodia, which make adhesive contacts with various substrates, such as other cells and the extracellular ma- trix. Differences in the adhesiveness of these contacts with different substrates influence the direction of mo- tility related events, but it is also through these adhesive contacts that growth cones can exert mechanical force on the substrate. Although the transmembrane linkage be- tween substrates and the cytoskeleton has been well stud- ied (reviewed in Burridge et al., 1988), the details of the interactions between substrates and cytoskeletal proteins remain unclear.

The major components of focal contacts are actin filament bundles and the actin-associated proteins alpha- actinin, vinculin, and talin. Alpha-actinin can bind di- rectly to and cross-link actin filaments (Bennett et al., 1984). It has also been shown to bind vinculin (Otto, 1983; Wilkins et al., 1983), which in turn binds with high affinity to talin (Burridge and Mangeat, 1984). The reported binding of talin to the transmembrane, cell ad- hesion protein integrin, however, occurs with relatively low affinity (Honvitz et al., 1986). It has recently been reported that alpha-actinin can bind directly to integrin (Otey et al., 1990). While these results suggest several viable models for the transmembrane linkage of the cy- toskeleton with extracellular substrates, it is possible that additional interactions and proteins are involved.

One suitable preparation in which to study these interactions is a homogeneous population of isolated, purified growth cones. While it has not proven feasible

Received March 19, 1991; revised April 29, 1991; accepted May 9, 1991. Address reprint requests to C. Cypher, Department of Cell Biology and Neuroanatomy, University of Minnesota, 4-135 Jackson Hall, 321 Church St. S.E., Minneapolis, MN 55455.

0 1991 Wiley-Liss, Inc.

260 Cypher and Letourneau

to obtain populations of growth cone particles from a single type of neuron, there are two published procedures for isolating fractions enriched in GCPs from developing brain (Pfenninger et al., 1983; Gordon-Weeks and Lock- erbie, 1984). Both procedures yielded subcellular frag- ments that were initially characterized as being enriched in growth cones by morphological criteria. In addition, radiolabeled growth cones microdissected from neuronal explant cultures copurified with GCPs from fetal rats (Pfenninger et al., 1983).

This initial characterization has been reinforced by demonstrations that these fractions are enriched in par- ticular proteins that have been shown to be concentrated in growth cones by immunofluorescence. For example, the 5B4 antigen (Wallis et al., 1985), growth associated protein-43 (GAP-43) (Meiri et al. , 1986, 1988), and the protooncogene product pp6Oc-src (Maness et al., 1988) have all been shown to be concentrated in growth cones by immunofluorescence, and all are enriched in isolated GCPs .

We have slightly modified the procedure of Pfen- ninger et al. (1983) to obtain a fraction from developing chick brain that is enriched in GCPs. Here we charac- terize this fraction with respect to its cytoskeletal pro- teins, proteins associated with focal adhesions and cell adhesion molecules. We report for the first time that isolated GCPs contain myosin 11, the high molecular weight neurofilament protein, and two proteins that in- teract with microtubules, kinesin and dynamin. In addi- tion, we find all of the major components of focal adhe- sions and several cell adhesion molecules. These results suggest that this preparation may be a useful system with which to study the interactions between these compo- nents.

MATERIALS AND METHODS Materials

Electrophoresis and blotting reagents were ob- tained from Bio-Rad (Richmond, CA) and ICN (Cleve- land, OH). The BCA protein assay reagent was obtained from Pierce (Rockford, IL). All other chemicals were from Sigma (St. Louis, MO). Controlled-pore glass (CPG03000, 300 nm mean pore size) was obtained from Electro-Nucleonics (Fairfield, NJ).

Antibodies Antibodies were generously provided from the fol-

lowing sources: monoclonal anti-GAP-43, Dr. P. Skene, Stanford University; rabbit anti-actin, Dr. J. Scholl- meyer, Roman L. Hruska Lab, NE; monoclonal antibod- ies against chick brain myosin 11, Dr. A.H. Conrad, Kansas State University; monoclonal antibodies to the high molecular weight neurofilament protein, Dr. G .

Shaw, University of Florida; monoclonal anti-beta-tu- bulin, Amersham (Arlington Heights, IL); rabbit anti- kinesin, Dr. J.R. McIntosh, University of Colorado; rab- bit anti-dynamin, Dr. R.L. Margolis, Fred Hutchinson Cancer Research Center, WA; rabbit anti-alpha-actinin and anti-talin, Dr. K. Burridge, University of North Carolina; anti-vinculin and anti-A-CAM, Dr. B. Geiger, Weizman Institute, Israel; antibodies against the beta 1 integrin subunit, Dr. S. Johansson, University of Upp- sala, Sweden, and Dr. A. Horwitz, University of Illinois; anti-L1, Dr. V. Lemmon, University of Pittsburgh; mouse anti-fibronectin, Dr. L. Furcht, University of Minnesota; anti-laminin, Dr. S. Palm, University of Minnesota. Secondary antibodies were from Promega (Madison, WI).

Growth Cone Particle Isolation GCPs were isolated after Pfenninger et al. (1983).

All procedures were done at 4°C. Whole brains from E13-El4 chick embryos were homogenized in six voP- umes of buffered sucrose (320 mM sucrose, 2 mM EDTA, 0.02% NaN, and 5 mM HEPES, pH 7.3) with 1 bg/ml aprotinin, 1 pM pepstatin, and 0.1 mM phenyl- methanesulfonyl fluoride. This crude homogenate was spun at 1,300g for 15 min. The supernatant from this low speed spin was layered over buffered sucrose (as above but with 750 mM sucrose) and spun at 150,OOOg for 60 min in a swinging bucket rotor. The material at the in- terface of this discontinuous sucrose gradient was re- moved and made 0.1 mM in phenylmethanesulfonyl flu- oride. This interface fraction was then chromatographed on a controlled-pore glass (CPG) column equilibrated with a modified Krebs’ buffer (145 mM NaCl, 5 mM KCl, 1.2 mM NaH,PO,, 1.2 mM MgCI,, 0.02% NaN,, and 5 mM HEPES, pH 7.3) (Gordon-Weeks and Lock- erbie, 1984). Fractions were monitored by absorbance at 280 nm. GCPs are found in the first peak off the column. When required, they were concentrated from pooled fractions by centrifugation at 84,OOOg for 30 min and resuspended in modified Krebs’ buffer.

Other Methods Protein concentrations were measured with the

Pierce BCA protein assay reagents using BSA as a stan- dard. SDS-PAGE (Laemmli, 1970) was performed on 7 or 7.5% gels. Gels were either stained according to Neu- hoff et al. (1 988) or blotted to nitrocellulose after Towbin et al. (1979) with 0.1% SDS in the transfer buffer. Blots were washed with, and antibody incubations performed in, Tris buffered saline (150 mM NaCl, 0.02% NaN,, and 10 mM Tris, pH 8.0) with 0.05% Tween 20. Alka- line phosphatase conjugated secondary antibodies were used according to the provider’s instructions.

Proteins in Chick Growth Cone Particles 261

Fig. 1. Isolated chick brain growth cone particles. GCPs from the first peak off the CPG column were allowed to settle onto polyornithine-coated coverslips and photographed by phase microscopy (500 X ) .

RESULTS We have isolated and begun to characterize a frac-

tion from embryonic chick brain that is enriched in GCPs. Following an existing procedure (Pfenninger et al., 1983; see also Gordon-Weeks and Lockerbie, 1984), with slight modifications, we consistently obtain a uni- form population of spherical, membrane-bound particles that are 1.5 to 2.0 pm in diameter. These are found in the first peak off the last step of the isolation procedure, chromatography on a CPG column. This preparation rou- tinely yields approximately 0.5 mg of GCPs per gram (wet weight) of embryonic chick brain. When allowed to settle onto polyornithine-coated coverslips, these parti- cles adhere firmly, flatten, and assume irregular shapes (Fig. 1).

An electrophoretic analysis of the polypeptide composition of all the steps in the isolation procedure is shown in Figure 2. In the gel shown here, GCPs from the first peak off the CPG column were concentrated by centrifugation for gel analysis. Under the conditions

Fig. 2. Polypeptide compositions of the steps in the isolation procedure. In this Coomassie stained gel the lanes are, from the left, (A) the crude homogenate, (B) the first supernatant, (C) the first pellet, (D) the supernatant, (E) interface and (F) pellet of the discontinuous sucrose gradient, (G) pelleted GCPs from the first peak off the CPG column, and (H) the second peak off the column. Each lane was loaded with the same amount of protein. Polypeptides enriched in the GCP fraction are indi- cated by arrowheads to the right of lane G. The molecular weight markers to the right are: 200, 116, 97, 66, and 43 kD.

used, 94% of the protein in this peak is pelleted, indi- cating that it is indeed found in a particulate fraction.

The Coomassie stained gel in Figure 2 reveals many polypeptides in all fractions, including the pelleted GCPs, where more than twenty bands are seen. Silver stained gels of GCP proteins reveal more than three times this number of bands (not shown). While some polypep- tides appear in all fractions, most notably the tubulins and actin (see below), it is clear that some polypeptides are enriched or depleted in various fractions. For exam- ple, compared to the crude brain homogenate the GCPs are enriched in polypeptides of approximately 187, 145, 89, 87, 75, and 28 kD. Compared to the interface frac- tion from the discontinuous sucrose gradient or the sec- ond peak off the CPG column, however, the GCPs are depleted in at least five polypeptides running above the myosin molecular weight marker, three bands of about 167 kD and bands of approximately 133, 124, and 92 kD. None of these polypeptides has been identified, al- though these differences in protein composition are con- sistently observed. Indeed the compositions of GCPs from separate isolations are always remarkably similar.

While there are no known unique protein markers for growth cones, there are proteins known to be en- riched in growth cones. For example, GAP-43 has been shown to be enriched in growth cones by immunoblots (Meiri et al., 1986) and immunofluorescence (Meiri et

262 Cypher and Letourneau

Fig. 3. GAP-43 immunoreactivity in the various purification steps. A gel of all fractions in the procedure, similar to Figure 2 , was transferred to nitrocellulose that was probed with a monoclonal antibody to GAP-43. The lanes are labeled as in Figure 2 . The following lanes are enriched in GAP-43: the sucrose interface fraction (lane E), the second pellet (lane F), and especially the GCPs (the pooled first CPG peak) (lane G). Each lane was loaded with the same amount of protein. The molecular weight markers to the right are: 200, 116, 97, 66, and 43 kD.

al., 1988). In order to determine if our preparation of GCPs was enriched in GAP-43, an unstained gel, con- taining all the fractions of the purification procedure, like that shown in Figure 2, was blotted onto nitrocellulose that was probed with a monoclonal antibody to GAP-43 (Fig. 3). While there is apparently proteolysis of GAP-43 at several steps in the procedure, it is clear that the iso- lated GCPs are greatly enriched in GAP-43.

We have further characterized these GCPs by qual- itatively assaying for three groups of proteins: cytoskel- eta1 proteins, proteins associated with focal adhesions, and cell adhesion molecules. Proteins within each of these groups were directly detected on immunoblots of GCP proteins that were probed with specific antibodies to the protein. In all cases the antibodies reacted with an antigen of the appropriate molecular weight.

The immunoblots for cytoskeletal proteins are shown in Figure 4. In agreement with many studies using immunofluorescence or electron microscopy (e.g., Le- tourneau, 198 1; Bridgman and Dailey, 1989; Letourneau and Shattuck, 1989), actin is a prominent component of GCPs, as is myosin 11. In addition, there is some neuro- filament protein present, as reflected by the presence of the high molecular weight neurofilament subunit. Tubu- lin is also present and densitometer scans of Coomassie stained gels show that the tubulins are the major compo- nent of this fraction (not shown). Kinesin (Vale, 1987)

Fig. 4. Cytoskeletal proteins in isolated growth cone particles. Pelleted GCP proteins were blotted onto nitrocellulose that was probed with antibodies to actin (lane A), myosin I1 (lane B), the high molecular weight neurofilament protein (lane C), beta-tubulin (lane D), kinesin (lane E), and dynamin (lane F). The molecular weight markers to the right are: 200, 116, 97 , 66, and 43 kD.

and dynamin (Shpetner and Vallee, 1989), two proteins that interact with microtubules, are also found in the isolated GCPs.

Because substrate adhesion is a major growth cone function, the GCPs were examined for several compo- nents of focal adhesions. In our preparations we can identify alpha-actinin, vinculin, talin, and the beta 1 sub- unit of integrin (Fig. 5) . These are all of the major com- ponents currently thought to be in focal adhesions (Bun- idge et al., 1988).

In addition to integrin, four other cell surface mol- ecules involved in cell adhesion are also present in the GCPs. They are A-CAM (equivalent or closely related to N-cadherin), L1 (equivalent to Ng-CAM), fibronectin, and laminin (Fig. 6). All of the antibodies used recognize antigens of the appropriate molecular weight. In the case of L1, the polyclonal antibody used recognizes both the 200 and 80 kD subunits (Rathjen and Schachner, 1984). In the case of laminin, the antibody recognized only the 205-215 kD beta subunits (Martin, 1987).

DISCUSSION We have modified slightly the procedure used by

Pfenninger et al. (1983) to isolate a fraction enriched in growth cone particles from developing chick brain. We have taken some additional care to limit proteolysis and have used a modified Krebs’ solution (Fried and Blaustein, 1978, cited in Gordon-Weeks and Lockerbie, 1984) for chromatography on the CPG column. This is the last step in the isolation procedure and separates the

Proteins in Chick Growth Cone Particles 263

Fig. 5. Components of focal contacts in isolated growth cone particles. Blotted proteins from GCPs were probed with anti- bodies to alpha-actinin (lane A), vinculin (lane B), talin (lane C), and the beta 1 subunit of integrin (lane D). The molecular weight markers to the right are: 200, 116, 97, and 66 kD.

Fig. 6. Cell adhesion molecules in isolated growth cones. Blotted proteins from isolated GCPs were probed with anti- bodies to A-CAM (equivalent or closely related to N-cadherin) (lane A), L1 (equivalent to Ng-CAM) (lane B), fibronectin (lane C), and laminin (lane D). The molecular weight markers to the right are: 200, 116, 97, 66, and 43 kD.

GCPs from soluble proteins as well as smaller particles. The protein compositions of the interface fraction from the sucrose gradient that was applied to the column, and the two peaks off the CPG column, are distinctly differ- ent (Fig. 2). This suggests that this is an important pu- rification step in the procedure. We use this column with the modified Krebs’ solution of Gordon-Weeks and Lockerbie (1984), rather than sucrose, to facilitate pel- leting the GCPs after chromatography.

We have used whole brains from chick embryos (E13-E14) rather than fetal or neonatal rat brain. Em- bryos of this age were chosen because by day 12 of chick neural development, cell proliferation has subsided in most areas of the brain while cell differentiation contin- ues (Romanoff, 1960). During this period, days 11-13, there is an approximately 50% increase in the amount of actin present in chick brain expressed as percent of total protein (Santerre and Rich, 1976). Microtubule protein, also expressed as percentage of total protein, reaches its

maximum concentration between days 9 and 17, when it is about double the percentage found in adult brain (Bamberg et al., 1973). While nerve fiber tracts start to form very early in chick development (Romanoff, 1960), these increases in the amounts of cytoskeletal proteins, after most cell division has been completed, probably reflect extensive neurite outgrowth during this period.

The GCPs obtained from chick brain appear, by phase microscopy, similar in size to those isolated from fetal rat brain (Pfenninger et al., 1983) or neonatal rat forebrain (Gordon-Weeks and Lockerbie, 1984) (Fig. 1). In addition, they adhere to and flatten out on coated coverslips as do those from rat forebrain (Gordon-Weeks and Lockerbie, 1984).

Protein profiles of the entire isolation procedure as presented here (Fig. 2), we believe, have not been re- ported. The major cytoskeletal proteins tubulin and actin appear to be present throughout the procedure, as one would expect. There are, however, distinct differences between some of the fractions, which will need to be examined in the future. It is difficult to compare the GCP protein profile obtained here with those from the two other isolation procedures without examining them on the same gel. They do appear to be similar. However, as the three preparations are derived from different sources and at different developmental stages, there may well be differences in protein composition.

Compared to the crude brain homogenate, the chick GCPs are greatly enriched in GAP-43 (Fig. 3). The antigen appears in Figure 3 to be a 57-kD polypeptide. GAP-43 has been assigned molecular weights of 43 to 57 kD, depending upon the electrophoretic conditions used (Skene, 1989). GAP-43 is a growth-associated protein whose expression is correlated with neurite growth dur- ing development and neuronal regeneration (Skene, 1989). It has been shown to be enriched in GCPs from neonatal rats by immunoblots and immunofluorescence (Meiri et al., 1986, 1988). Although the function of GAP-43 is not known, expression of GAP-43 in several cell types causes the cells to extend long filopodial-like processes (Zuber et al., 1989). While GAP-43 is appar- ently absent from dendritic growth cones (Goslin et al., 1990), it may still be important for more rapid axonal growth.

The chick GCPs possessed all of the cytoskeletal proteins examined (Fig. 4). In agreement with many other studies on intact and isolated growth cones, actin, myosin 11, and tubulin are all present (e.g., Letourneau, 1981; Gordon-Weeks, 1987; Bridgman and Dailey, 1989; Simkowitz et al., 1989). In intact growth cones, actin filament bundles are found in the filopodia and as a dense network in the lamellipodia (Bridgmen and Dai- ley, 1989). Anti-myosin I1 immunofluorescence appears punctate and is found at the bases of filopodia and the

264 Cypher and Letourneau

proximal edge of the thinly spread peripheral region (Bridgman and Dailey, 1989). The exact roles of actin and myosin I1 in growth cone motility are still not clear. In this regard, it is noteworthy that we do have prelim- inary immunoblot data, using a monoclonal antibody elicited by chicken brush border myosin I, that myosin I is greatly enriched in the GCPs compared to the crude homogenate (Cypher and Mooseker, unpublished obser- vation). The role of myosin I in growth cone motility also needs to be elucidated.

We report here for the first time the presence of neurofilament protein in GCPs (Fig. 4). Neurofilaments are not usually thought to be components of the growth cone cytoskeleton. They are not seen as formed filaments in growth cones and are thought to form proximal to the more dynamic actin and microtubule cytoskeleton, add- ing stability to the forming neurite (Shaw, 1987). The high molecular weight neurofilament protein identified here may be a contaminant in the GCP preparation from the distal regions of neurites.

The presence of kinesin and dynamin in GCPs is reported here for the first time (Fig. 4). The role of kinesin in microtubule-associated motility and the trans- port of membrane-bound vesicles has been well studied (reviewed in Vale, 1987). In intact growth cones micro- tubules are found centrally, splaying out of the neurite (e.g., Bridgman and Dailey, 1989). Some individual mi- crotubules, extending from the central region, reach as far forward as the bases of individual filopodia (Le- tourneau and Ressler, 1983). Kinesin may have an im- portant role in the transport of materials to and within the growth cone during neurite elongation. Less is known about dynamin (Shpetner and Vallee, 1989), which me- diates the ATP-sensitive bundling of microtubules and may associate with membranes as well. Dynamin has been localized to the cell bodies and growth cones of PC12 cells by immunofluorescence (Scaife and Mar- golis, 1990). We have made similar observations in pri- mary cultures of dorsal root ganglion neurons (Cypher and Letourneau, unpublished observations).

The adhesive properties of isolated GCPs have not been extensively studied. The chick GCPs will adhere to and flatten out on polyornithine-coated coverslips. Gor- don-Weeks and Lockerbie (1984) reported similar flat- tening of rat GCPs on poly-D-lysine-coated coverslips. Lockerbie et al. (1989) have reported that rat GCPs ad- here poorly to untreated plastic, and only somewhat bet- ter to poly-L-ornithine-treated plastic. However, up to 18% of labeled GCPs were observed to adhere to a trans- formed, mouse glial cell line. This adhesion was inhib- ited by pretreatment of the GCPs with trypsin and con- ditions that would elevate CAMP levels, suggesting that the adhesion was dependent on particle surface proteins and under regulation.

We have found that isolated chick GCPs contain all of the known major components of focal adhesions; al- pha-actinin, vinculin, talin, and integrin (Fig. 5). In ad- dition to integrin, they contain the cell adhesion mole- cules A-CAM, L1, fibronectin, and laminin (Fig. 6). A current model of focal contacts (Burridge et al., 1988) proposes that cytoskeletal actin filaments, cross-linked by alpha-actinin, are linked via vinculin and talin to transmembrane integrin receptors that bind to extracel- lular matrix molecules such as fibronectin and laminin. The results reported here identify all of these components in isolated GCPs.

The several cell adhesion molecules we have iden- tified in this preparation suggest that there may be others. It must be remembered that these GCPs were derived from whole brain at a particular period in chick neural development. The cell adhesion molecules present in GCP preparations may vary with the age of the embryos used and the regions of brain sampled. Taken together, the results presented here suggest that this preparation of chick GCPs may be a useful model system with which to study growth cone adhesion.

ACKNOWLEDGMENTS We would like to thank all those who provided

antibodies for this study, Dr. Karen Mesce for comments on the manuscript, and Jerry Sedgewick for photographic help. This research was supported by NIH grant HD19950.

REFERENCES Bamburg JR, Shooter EM, Wilson L (1973): Developmental changes

in microtubule protein of chick brain. Biochemistry 12: 1476- 1482.

Bennett JP, Zaner KS, Stossel TP (1984): Isolation and some proper- ties of macrophage alpha-actinin: Evidence that it is not an actin gelling protein. Biochemistry 235081-5086.

Bridgman PC, Dailey ME (1989): The organization of myosin and actin in rapid frozen nerve growth cones. J Cell Biol 108:

Bumdge K , Fath K , Kelly T, Nuckolls G (1988): Focal adhesions: Transmembrane junctions between the extracellular matrix and the cytoskeleton. Ann Rev Cell Biol 4:487-525.

Bumdge K, Mangeat P (1984): An interaction between vinculin and talin. Nature 308:744-746.

Fried RC, Blaustein MP (1978): Retrieval and recycling of synaptic vesicle membrane in pinched-off nerve terminals (synapto- somes). J Cell Biol 78:685-700.

Gordon-Weeks PR (1987): The cytoskeletons of isolated, neuronal growth cones. Neuroscience 21:977-989.

Gordon-Weeks PR, Lockerbie RO (1984): Isolation and partial char- acterization of neuronal growth cones from neonatal rat fore- brain. Neuroscience 13:119-136.

Goslin K, Schreyer DJ, Skene JHP, Banker G (1990): Changes in thlz distribution of GAP-43 during the development of neuronal polarity. J Neurosci 10588-602.

95-109.

Proteins in Chick Growth Cone Particles 265

Otto JJ (1983): Detection of vinculin-binding proteins with an Iz5I- vinculin gel overlay technique. J Cell Biol 97: 1283-1287.

Pfenninger KH, Ellis L, Johnson MP, Friedman LB, Somlo S (1983): Nerve growth cones isolated from fetal rat brain: Subcellular fractionation and characterization. Cell 35573 -584.

Rathjen FG, Schachner M (1984): Immunocytological and biochem- ical characterization of a new neuronal cell surface component (L1 antigen) which is involved in cell adhesion. EMBO J 3: 1-10,

Romanoff AL (1960): “The Avian Embryo.” New York: Macmillan c o .

Santerre RF, Rich A (1976): Actin accumulation in developing chick brain and other tissues. Dev Biol 54:l-12.

Scaife R, Margolis RL (1990): Biochemical and immunochemical analysis of rat brain dynamin interaction with microtubules and organelles in vivo and in vitro. J Cell Biol 111:3023-3033.

Shaw G (1987): Neurofilaments: Abundant but mysterious neuronal structures. BioEssays 4: 161-166.

Shpetner HS, Vallee RB (1989): Identification of dynamin, a novel mechanochemical enzyme that mediates interactions between microtubules. Cell 59:421-432.

Simkowitz P, Ellis L, Pfenninger KH (1989): Membrane proteins of the nerve growth cone and their developmental regulation. J Neurosci 9: 1004-101 7.

Skene JHP (1989): Axonal growth-associated proteins. Ann Rev Neu- rosci 12: 127-156.

Towbin H, Staehelin T, Gordon J (1979): Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Pro- cedure and some applications. Proc Natl Acad Sci USA 76: 4350-4354.

Vale RD (1 987): Intracellular transport using microtubule-based mo- tors. Ann Rev Cell Biol 3:347-378.

Wallis I, Ellis L, Suh K, Pfenninger KH (1985): Immunolocalization of a neuronal growth-dependent membrane glycoprotein. J Cell Biol 10 1 : 1990-1998.

Wilkins JA, Chen KY, Lin S (1983): Detection of high molecular weight vinculin binding proteins in muscle and nonmuscle tis- sues with an electroblot-overlay technique. Biochem Biophys Res Comm 116:1026-1032.

Zuber MX, Goodman DW, Karns LR, Fishman MC (1989): The neuronal growth-associated protein GAP-43 induces filopodia in non-neuronal cells. Science 244:1193-1195.

Honvitz A, Duggan K, Buck C, Beckerle MC, Burridge K (1986): Interactions of plasma membrane fibronectin receptor with talin-a transmembrane linkage. Nature 320531-533.

Laemmli UK (1970): Cleavage of structural proteins during assembly of the head of bacteriophage T4. Nature 227:680-685.

Letourneau PC (198 1): Immunocytochemical evidence for colocaliza- tion in neurite growth cones of actin and myosin and their relationship to cell-substratum adhesion. Dev Biol 85:

Letourneau PC, Ressler AH (1983): Differences in the organization of actin in the growth cones compared with the neurites of cul- tured neurons from chick embryos. J Cell Biol 97:963-973.

Letourneau PC, Shattuck TA (1989): Distribution and possible inter- actions of actin-associated proteins and cell adhesion molecules of nerve growth cones. Development 105505-519.

Lockerbie RO, Autillo-Touati A, Araud D, Seite R, Chneiweiss H, Glowinski J, Prochiantz A (1989): Cyclic AMP reduces adhe- sion of isolated neuronal growth cones from developing rat forebrain to an astrocytic cell line from embryonic mouse stri- atum. Neuroscience 28443-454.

Maness PF, Aubry M, Shores CG, Frame L, Pfenninger KH (1988): c-src gene product in developing rat brain is enriched in nerve growth cone membranes. Proc Natl Acad Sci USA 85:5001- 5005.

Martin GR (1987): Laminin and other basement membrane compo- nents. Ann Rev Cell Biol 357-85.

Meiri KF, Pfenninger KH, Willard MB (1986): Growth-associated protein, GAP-43, a polypeptide that is induced when neurons extend axons, is a component of growth cones and corresponds to pp46, a major polypeptide of a subcellular fraction enriched in growth cones. Proc Natl Acad Sci USA 83:3537-3541.

Meiri KF, Willard M, Johnson MI (1988): Distribution and phosphor- ylation of the growth-associated protein GAP-43 in regenerat- ing sympathetic neurons in culture. J Neurosci 8:2571-2581.

Neuhoff V, Arold N, Taube D, Ehrhardt W (1988): Improved staining of proteins in polyacrylamide gels including isoelectric focus- ing gels with clear background at nanogram sensitivity using Coomassie Brilliant Blue G-250 and R250. Electrophoresis 9:

Otey CA, Pavalko FM, Burridge K (1990): An interaction between alpha-actinin and the Beta-1 integrin subunit in vitro. J Cell Biol 111:721-729.

113-122.

255 -262.