Molecular mechanisms of axon growth and guidance.

ANNUAL REVIEWS

Annu. Rev. Cell Bioi. 1991. 7: 1/7-59 Copyright © 1991 by Annual Reviews Tnc. All rights reserved

Further

Quick links to online content

MOLECULAR MECHANISMS OF Annu. Rev. Cell. Biol. 1991.7:117-159. Downloaded from www.annualreviews.org by Pennsylvania State University on 08/02/12. For personal use only.

AXON GROWTH AND GUIDANCE John L. Bixby Department of Molecular and Cellular Pharmacology, University of Miami, School of Medicine, Miami, Florida 3310 I

William A. Harris Department of Biology, University of California, San Diego, La Jolla, California 92093 KEY WORDS:

growth cone, neurite outgrowth, growth promoting, axonal guid ance, ECM, CAMs, neurotrophic, neurotropic, adhesion, trans duction, in vitro systems

CONTENTS INTRODUCTION ............................................................................................................. .

118

PROTEINS THAT PROMOTE NEURITE OUTGROWTH ........................................................... .

119 122 123 124 126

Soluble Neurotrophic Agents . . . . . . . . .. ..... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. ... . . . . . . . . . . . .. . . . . .......... . . . ECM Proteins . . . ........... . . . . . . . . . ............. . . . . . . . . . . .............. . . . . . . .. . . . . . . . . . . . . . . . . .. . . . .. . ... . . . . . . . . . . CAMs Persp ectives .

. .

...................................................................................................................... .........

RECEPTORS FOR THE GROWTH PROMOTER8.................................................................... . .

Neurotrophic Receptors . . . . . . . . . . . .. . . .. . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . ............ . . . . . . . . .. . .. . . . . . . . . . . . . . . . . . . . ECM Receptors . . . . . . . . . . . . . .. . . .. . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . ............. CAM Receptors . . . ........... . . . . . . .. . . . . . . . . . . . . . . . . . . .. . . . . . . . . . .. . . . . . . . . . .. . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . Perspectives ........................... . . . . . . . . ............. . . . . .. ............ . . . . . . . . ................................... . .

.

.

SECOND MESSENGERS IMPLICATED IN NEURITE GROWTH .................................................. .

Calcium ................... . . ................. . . . . . . ........ ...... . . . . . ....... . . . ................... . . . . . . ........... . . . . . . cAMP ............................................. . . . . . . . . .. . . . . . . . . . . . . . . .. .............................................. G Proteins and ras ................................................................................................... Tyrosine Kinases and src ......................................................................................... Protein Kinase C . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . ....... . . . . . . .. . . . . . . . . . ...... . . . . . . . . . . Nuclear Events ........................................................................................................ Perspectives . . ........ . . . . . . . . . . . . . . . . . . . . . .. . . . . . .. . . . . . . . . . . . . . . . . .. . . . . . . . . . . ........ . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . .

127 127 128 128 1 29 129 130 130 131 132 133 1 34 1 34 117

0743--4634/91/1115-0117$02.00

1 18

BIXBY & HARRIS

MOLECULES THAT GUIDE AXONS . . . . . . . . . . . . . ...................................... . .. . . . .. . . . . . . . . . . . . . . . . . . . . . . .. ..

Guidance vs Growth Promotion . . . . . . ...... . .................................. . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... Cell Surface Proteins ............... .......... ...................................................................... ECM Molecules ................................ . . . . . . . . . . . . . ....... . . . . .............................................. Diffusible Neurotropic Factors.......... . . . . . . . . . . . . . . . . . . .... . . . . . . . ........................................... Inhibitory Factors............................................ . . . . . . . . . . . . . . . . . . ..... . .. . . . . . . . . . . . . . . . . . .. . ........... Topographic Guidance Molecules ......... . . . . . . . . . . . . . . . . . .................................................. Perspectives . .. . . . . . . . . . . . . . . . . . . . . . . .......................................................................... . . . . . . . . ...

134 134 138 141 142 144 145 148

CONCLUSIONS................................................................................................................

149

.

.

.

Annu. Rev. Cell. Biol. 1991.7:117-159. Downloaded from www.annualreviews.org by Pennsylvania State University on 08/02/12. For personal use only.

INTRODUCTION The mechanisms of axonal growth and guidance have been under inves tigation for over a century, since Cajal first discovered the growth cone. It was not until the 1 980s, however, that much headway was made into the molecular basis of outgrowth and guidance. Since then a multitude of factors appearing on cell surfaces, in the extracellular matrix, and diffusing through the brain have been described. Many of these factors promote axonal outgrowth, and many appear to be involved in growth cone guid ance. Here, we attempt to review the status and mechanisms of these factors. We are aware that we have not covered all the putative molecular components that have been implicated in growth promotion and guidance. For example, we do not address, due to our focus on the external factors that promote outgrowth of neurites and guide axons, the possible roles of proteases secreted by the growth cones in neurite advancement and penetration into the extracellular matrix (see Pittman et aI 1 989). We hope, however, that our effort will provide at least a framework for organizing many of the known growth and guidance factors and for cataloging newly discovered ones. Early opinions that axonal growth was regulated by gradients of chemo tropic molecules, by a plethora of labeled pathways, or by expression sequences of one or two cell adhesion molecules (CAMs) have given way to the view that each of these classes of control is exercised, and that the pattern of axonal growth that is found is the result of concerted action by numerous factors. In particular, three major classes of protein have been found to promote and guide neurite growth in vitro and/or axonal growth in vivo: soluble trophic factors and chemotropic agents, various con stituents of the extracellular matrix (ECM), and several different categories of cell surface molecules. Recent reviews have discussed these molecules in some detail (Berg 1 984; Lander 1 987; Jessell 1 988; Reichardt et al 1 989). In this review we shall use the term "neurite outgrowth-promoting molecules" to ·refer to molecules encountered by the neurons in their environment, rather than to neuronal cell surface receptors, which are dealt with separately.

AXONAL GROWTH AND GUIDANCE

1 19


PROTEINS THAT PROMOTE NEURITE GROWTH One of the emerging themes in the study of neurite growth-promoting factors is that regulation of growth involves multiple molecular mech anisms acting in concert. The first suggestion for this idea came from studies using antibodies to block the function of growth-promoting cell adhesion molecules (CAMs) and ECM proteins (Tomaselli et al 1986, 1 988; Chang et a1 1 987; Bixby et a1 1 987, 1 988). These studies demonstrated that several molecular mechanisms, involving both ECM proteins and different types of CAMs, exist on the same cell and can promote growth from the same neuron. For example, ECM proteins, the Ca 2 + -independent neural CAM, (NCAM), and the Ca 2 + -dependent CAM, N-cadherin, can function to promote neuronal growth on the surfaces of myotubes (Bixby et al 1987). Later similar conclusions from molecular genetic studies in Drosophila showed that mutation of a gene coding for a single adhesion molecule had a small effect on development (Zinn et al 1 988; Bieber et al 1 989; see also Gertler et al 1 989; Elkins et al 1 990). One interpretation of these results is that there is redundancy in molecular mechanisms of axon growth, presumably to guard against gross defects caused by a single mutation. Such an interpretation is supported by the results in flies (Bieber et a1 1989; Elkins et aI1 990). It is also possible, however, that these multiple mechanisms represent not simply redundancy, but a combinatorial process, such that the desired growth is achieved only by the concerted function of several molecule-receptor systems. This idea is consistent with the antibody results mentioned above and is also supported by the obser vation that the function of the CAM Ll is augmented by binding to NCAM in vitro (Kadmon et al 1 990). It has been suggested that one function of NCAM may be to modulate interactions involving many different adhesion molecules (Rutishauser et al 1 988). Synergism among growth-promoting molecules need not be limited to those that are sub strate-bound. For example, laminin (LN) and nerve growth factor (NGF) may be able to synergize in the induction of sympathetic neurites (Toma selli et al 1 988). The individual proteins that have been implicated in axon growth are listed in Table I . In developing a list of neurite growth-promoting proteins, several difficulties arise, some of which are due to our lack of sufficient knowledge, and some of which are due to problems with definitions. For example, the soluble growth-promoting agents listed in Table 1 have generally been demonstrated to have many different actions on the same cell. In many assays, it is difficult to separate the induction of neuronal differentiation or the enhancement of neuronal survival from the induction of neurite growth per se. Similarly, ECM glycoproteins and CAMs often

......


tv o

0:1

� Table I

0:1 -
j :;>j 1;:]


Ig superfamily NCAM MAG

1 80, 140 , 120 70

Ll, NILE, G4, NgCAM . 8D9 Contactin, FII

190, 135, 80 l 30

0ligodendrocytes, Schwann cells Nerve fibers, some glia Subsets of axons

135

Restricted in CNS

125

CNS, peripheral neurons, muscle, lens, heart Subsets of axons Restricted in CNS

TAG-I Others N-cadherin Neurofascin P84 P30

1 85, 160 1 67, 85, 66 30

CNS, PNS, muscle, glia

Brain (neurons?)

Some forms GPI-linked; fasciclin II is homologue

Neuroglian is homologue family of related proteins F3 is homologue, some forms GPI-linked GPI-linked; floor plate expression

Bixby et al ( 1987) D oherty et al ( 1990) Johnson et al ( 1989) Salzer et al ( 1987) Lagenauer & Lemmon (I 987) Chang e t al ( 1987) Ranscht ( 1988) Brummendorf et al ( 1989) Furley e t al (I 990)

Distribution like L l Floor plate expression

Takeichi ( 1 990) Bixby & Zhang (1990) Rathjen et al (1987) Chuang & Lagenauer ( 1 990)

Related to H MG-l DNA-binding protein

Rauvala et al ( 1988) Rauvala & Pihlaskari ( 1987)

Large family of cadherins

�Z �

Cl

�

� �

tj Cl e

a

�

n m

-

N

1 22

BIXBY

&

HARRIS

have potent adhesion-promoting activity, and many workers have failed to distinguish between this effect and the induction of neurite growth. For the proteins listed in Table I , evidence suggests a role in neurite growth that is independent of effects on survival or cell adhesion. We now discuss these molecules individually. Soluble Neurotrophic Agents

The soluble growth-promoting factors so far identified are called neurotrophic molecules because of their many and various effects on the survival and differentiation of neurons (e.g. Berg 1 984). The best characterized of these, nerve growth factor (NGF), was in fact the first molecule described that clearly promoted axon growth (Greene & Shooter 1 980). Studies of adult neurons, which do not require NGF for survival but remain responsive, show that NGF can promote growth of differentiated neurons despite being dispensable (Lindsay 1988). NGF, however, may not be involved in the initial axonal growth of responsive cells (Davies et al 1 987; Large et al 1 986). Recently it has become clear that NGF is a member of a family of growth-promoting molecules, and two other mem bers, brain-derived neurotrophic factor (BDNF) and neurotrophin-3 (NT3), have been identified and characterized (Barde 1 988; Leibrock et a1 1 989; Maisonpierre et al 1 990; Rosenthal et al 1 990). The three proteins show 5 0-60% sequence identity in the portion of the precursor molecule believed to be the active factor. NGF, BDNF, and NT-3 have similar but unique temporal and spatial distributions and are apparently active on distinct but overlapping populations of neurons (Barde 1 988; Masonpierre et al 1 990; Rosenthal et al 1 990). There is reason to believe that other members of this family exist (Rosenthal et al 1 990).


NGF, BDNF, NT3

Fibroblast growth factor (FGF) is a member of a family of heparin binding growth factors with a wide range of specificities and activities, both in the nervous system and elsewhere (Burgess & Maciag 1 989). Although it has no signal sequence and its means of secretion is con troversial (Abraham et aI 1 986), it can clearly be found in the extracellular milieu, including the ECM (Rogelj et al 1 989), and is a potent effector of neuronal survival and neurite outgrowth in vitro (Unsicker et al 1 987; Hatten et al 1 988). In some cases the effects of FGF on neurite growth can be separated from those on neuronal survival (Hatten et al 1 988).

FGF

SIOO{1 S I 00[3 is a glial-derived protein that is a member of a family of small acidic proteins (Van Eldik et al 1 988). Its activity was originally defined as neurite extension factor, or NEF (Kligman 1 982), and this factor was subsequently shown to consist of homodimers of S l 00[3 (Kligman & Mar shak 1 985; see also Winningham-Major et a1 1 989). Although this protein


1 23

is potent and can initiate a program of neurite growth with a short exposure to the neurons (Winningham-Major et al 1 989), it is not certain whether it is a true neurite induction factor or an inducer of neuronal differentiation. Insulin-like growth factor II (IGF-II) is a protein closely related to insulin (Rinderknecht & Humbel 1 978), found in serum and in the CNS, that has a number of trophic effects on some populations of neurons. IGF II can promote neuronal growth in a manner that seems to be independent of trophic effects, although this issue is not fully resolved (Recio-Pinto et al 1 986).


IGF-II

ECM Proteins

Several of the most abundant components of the ECM including laminin (LN), fibronectin (FN), and the collagens are capable of promoting neurite growth in vitro (for reviews see Rogers et al 1 989; Lander 1 987; Reichardt et aI 1989). These proteins are multifunctional and clearly cap able of promoting both cell adhesion and various aspects of cellular differ entiation in the nervous system and elsewhere (see Ruoslahti & Piersch bacher 1987, for examples). Nevertheless, the ability of ECM proteins to stimulate neurite growth can be separated from cell adhesion (e.g. Toma selli et al 1 986). Although the responsiveness of different neuronal types to different ECM proteins varies, all three of these major ECM components can promote growth from several different types of neurons (Rogers et al 1 989; Lander 1987). This, in combination with the broad distribution of the ECM proteins, has led to the view that no specific growth function can be attributed to them. One reason for thinking this conclusion premature is that all three proteins exist in multiple forms and, in the case of LN and FN, a single form can clearly have many different functional domains, possibly with distinct receptors. Examination of the gene for FN reveals that close to a dozen distinct polypeptides can be produced (Kornblihtt et al 1 98 5), and a single form of FN can have multiple neurite-promoting domains (Humphries et al 1 988; Rogers et al 1989). Similarly, LN can exist in various combinations of A, B I , and B2 chains (Edgar et al 1 988; Klein et al 1988), and several different functional domains with different receptors have been described (e.g. Sonnenberg et al 1 990). It is also likely that the association of LN with other proteins and proteoglycans alters its functional state (Chiu et al \986). To add to the complexity, it has recently become clear that LN is not an individual molecule, but part of a family of proteins with at least two other members, s-LN (Hunter et al 1 989) and merosin (Ehrig et al 1990). Both proteins have restricted and unique distributions, and merosin can stimulate neurite growth (Manthorpe et al 1990). It seems likely that additional members of the LN family will be discovered. LN, FN

1 24

BIXBY & HARRIS

Another component of the ECM, thrombospondin (TSP), has been implicated in the control of neurite growth. TSP, like LN and FN, is a multifunctional glycoprotein that can mediate cell adhesion and spreading of a variety of cell types (Lawler & Hynes 1 986; Taraboletti et al 1 987). Its distribution during development is consistent with a role in neurite growth (O'Shea & Dixit 1 988), and it was recently shown to directly promote axonal growth in vitro (Osterhout & Higgins 1 990; Neugebauer et al 1 99 1). Various collagens have limited neurite outgrowth activity in vitro, although some forms may be excellent growth substrates for subpopulations of neurons (e.g. Kleitman et al 1 988). It is possible that these proteins act in conjunction with other ECM components.


TSP, COLLAGENS

CAMs NCAM, MAG Many of the proteins now recognized as axonal growth fac tors are CAMs, and many of these are members of the immunoglobulin (Ig) superfamily. In fact, the regulation of cell adhesion probably represents the original function of Jg superfamily proteins (Williams & Barclay 1 988). These molecules can be either integral membrane proteins, or linked to the membrane through glycophosphatidylinositol (GPI). The most extensively characterized of these CAMs, NCAM (Edelman 1 986), has a demonstrated ability to promote neurite growth. This has been shown with antibody blocking studies (Bixby et al 1 987) and through transfection of the gene for NCAM into heterologous cells (Doherty et al 1 989, 1 990), although the purified protein is relatively inactive in most assays (e.g. Lagenauer & Lemmon 1987; Bixby & Jhabvala 1 990). Besides its relatedness to Ig superfamily molecules, NCAM has several repeat units homologous to the type III domains of FN (Cunningham et aI 1 987). The apparent homologue of NCAM in Drosophila, fasciclin II, was discovered by a search for axon growth-promoting molecules, and it may serve such a function (Harrelson & Goodman 1 988). A CAM with significant homology to NCAM, the myelin-associated glycoprotein (MAG), is traditionally thought of as being involved in myelin formation (see Salzer et al 1 987), but this protein has also been implicated in neurite growth through cDNA transfection studies (Johnson et aI1 989). 1t should be noted that Thy-I, a simple Ig superfamily member related to NCAM, has been implicated in neurite outgrowth as well (Leifer et al 1 984). L I, NgCAM, FlljCONTACTIN, TAG·l, NEUROFASCIN A second group of Ig super family molecules comprises L l , NgCAM , F I I /contactin, and TAG- I , although other members of this group seem certain to be described. A fifth member of this group, called NgCAM-reiated molecule (Nr), has recently been cloned (M. Grumet et aI, submitted). Nr is 40% identical both to



1 25

NgCAM and to L l , which themselves are only 40% identical (Burgoon et al 1 99 1 ) . The members of this group of proteins are all similar in primary sequence (30-50%) as well as overall structure, with each con taining six Ig-type repeat units and four or five FN-type III domains (Moos et al 1 988; Brummendorf et a1 1 989; Ranscht 1 988; Furley et aI1990). On the basis of sequence identity, it is possible to group L l (NILE), NgCAM , and N r in one subfamily, with TAG- 1 and F l ljcontactin in another. A member of this family has also been described in Drosophila, called neuroglian (Bieber et al 1 989), which may be the Drosophila homologue of L l . Another CAM, neurofascin (Rathjen et al 1 987), shows a similar pattern of expression to Ll and F l l/contactin (widespread distribution on neuronal fibers) and may be a member of this family, although its primary structure is unknown. In contrast to the broad distribution of these three, TAG- 1 shows a greatly restricted distribution during embryo genesis and may be involved in guidance functions of the floor plate (see below). L l , which has been identified a number of times in different species, is likely to be the same molecule as NILE, G4, and 8D9 (see Lemmon et al 1 989, for discussion), although the situation may be more complex (Burgoon et al 1 99 1 ). The role of L l in the induction of neurite growth has been demonstrated through the use of the purified protein as a substrate (e.g. Lagenauer & Lemmon 1 987) and by antibody blocking studies (Chang et a1 1 987; Bixby et a1 1 988; Seilheimer & Schachner 1 988). F l l /contactin and neurofascin have been implicated in the induction of axon growth by antibody blocking experiments, and purified TAG- l has been shown directly to promote neurite growth from some (but not all) neurons (Chang et al 1 987; Rathjen et al 1 987; Fudey et al 1 990). The mouse homologue of F 1 1/contactin is the protein F3 (Gennarini et al 1 989). The members of the Ig superfamily mentioned above belong to the general class of Ca 2+ -independent CAMs. A second major group of CAMs is the Ca2+ -dependent CAMs, or cadherins (Takeichi 1 988), of which five or six members have clearly been identified, and others of which certainly exist (Takeichi 1 990; Crittenden et al 1 988; Ranscht & Dours 1 989; c. Kintner, personal communication). The major cadherin in the nervous system appears to be N-cadherin, probably identical to molecules originally identified as N-caICAM and A-CAM (see Bixby & Zhang 1 990, for references). N-cadherin has been implicated in neurite growth by all three of the criteria currently in use: inhibition of neurite growth by antibodies to the protein (Bixby et al 1 987, 1988; Tomaselli et al 1 988; Neugebauer et al 1 988), enhancement of the neurite-promoting activity of cells transfected with the N-cadherin cDNA (Matsunaga et al 1 988), and CADHERINS

1 26

BIXBY & HARRIS

induction of neurite growth using the purified protein (Bixby & Zhang 1 990). Other cadherins are likely to be present in the regions through which axons must navigate, and they might also influence growth, either in a positive or a negative fashion (Ranscht & Dours 1 989; M. Takeichi, per sonal communication). Other proteins listed in Table I are apparently unrelated to those mentioned above. P84 is a CAM with a restricted distribution in the embryonic CNS, like that of TAG- I , which later becomes more widely expressed (Chuang & Lagenauer 1 990). Like TAG- I , it is expressed in the floor plate region, and promotes neurite growth from some but not all neurons when used as a purified substrate. Preliminary analysis suggests that P84 is not identical to any of the previously characterized CAMs (c. Lagenauer, personal communication), but it remains to be seen whether it may belong to one of the families identified above. P30, a brain protein bound to membranes in a heparin-dependent fashion, promotes neurite growth from brain neurons (Rauvala & Pihlaskari 1 987; Rauvala et al 1 988). This molecule is widespread in late embryonic rat brain, but is down-regulated shortly after birth. Sequence analysis demonstrates that this protein is identical to the DNA-binding protein HMG- l . Which of the two suggested properties is primary remains unclear (Rauvala et al 1 988). Finally the secreted protein, axonin-I, which also has a membrane associated form (Ruegg 1 9 89), has recently been shown to act as an adhesive and possibly a neurite outgrowth-promoting molecule when puri fied (Stoekli et al 1 99 1).


P84, P30

Perspectives

How many molecules exist that promote axonal growth? In the past, views on the number varied considerably, from two or three, whose functions are regulated by modulation of their abundance and structure (Edelman 1 986), to more than 50 (Bastiani et al 198 5). Clearly, this question is unanswerable at present, but several points can be made. First, the number proposed will surely vary depending on one's definition of a growth promoting molecule. Second, the number of such molecules may not be astronomical because many of those noted in Table I have been isolated multiple times, in some cases using quite different strategies. Third, one of the themes that emerges from a perusal of Table 1 is that many of the proteins listed are members of families, including the LNs, the NGF related molecules, the Ig superfamily molecules, and the cadherins. In several cases, it is clear that different members of these families have subtly different distributions and/or functional properties. On evolutionary grounds, one might predict that different members of a family of related



1 27

factors will be used on distinct but evolutionarily related pathways. Whether or not this idea has any merit will surely be uncovered as students of neural evolution turn their attention to these molecules. The actual number of growth-promoting molecules, whatever it turns out to be, will have a much higher functional complexity than indicated by this base number if appropriate consideration is given to (a) temporal and spatial modulation of the expression of these proteins, alone and in com binatorial association, (b) similar regulation of their neuronal receptors, and (c) the potential for functional changes imposed on an individual molecular species by alternative splicing and post-translational modifications.

RECEPTORS FOR THE GROWTH PROMOTERS The growth-promoting molecules discussed above must influence axon elongation through binding to specific cell-surface receptors, which in turn must transduce this signal to the neuron's interior. Early hypotheses about the mechanism of this transduction focused on neuron-substrate adhesion, and the specific hypothesis developed that growth promotion (and guidance) by molecules like LN and FN was due to their ability to increase this adhesion (Letourneau 1 975; Hammarback et aI1988). The studies cited (and others) clearly demonstrated the necessity for adequate adhesiveness if neurites were to form. Now the strict adhesion hypothesis seems unlikely to be correct on the basis of several lines of evidence. (a) For physiologically relevant substrates, there is little or no correlation between the strength of neuron-substrate or growth cone-substrate adhesion and the ability of the substrate to induce process outgrowth (Gunderson 1 987; Hall et a1 1 987; Lemmon et aI 1 990). (b) Antibodies to neuronal receptors for these proteins can inhibit neurite growth even in conditions in which strong adhesion to the substrate and lamellipodial formation is allowed (e.g. Tomaselli et al 1986). (c) Antibodies to some, but not all, neuronal cell surface receptors can promote neurite growth when used as culture substrates (Leifer et al 1 984; Hall et al 1 987). The data therefore suggest that the involved recep tors transduce growth signals through intracellular messenger systems and the cytoskeleton, j ust as is the case for other cell surface receptors. What are the receptors involved in transduction of the signals provided by the proteins in Table I ? Neurotrophic Receptors

In the case of NGF, a 75-kd receptor (NGF-R) has been purified and cloned, and a great deal is known about its synthesis and regulation (for review, see Springer 1988). Although two affinity forms of the NGF-R

NGF·R


1 28

BIXBY & HARRIS

have been described, it seems likely that only one receptor protein exists and that its function is modulated by association with other proteins (Hempstead et al 1 989). The NGF-R is not related to any other family of cell surface receptors, so the transduction mechanism associated with it cannot be inferred by analogy. Indeed, the key to this transduction may lie in the proteins with which NGF-R associates. The structures of the receptors for BDNF and NTF-3 have not been described, but we may guess that they will prove to have some functions in common with the NGF-R. The recent finding that a receptor-class tyrosine kinase is acti vated by NGF suggests the possibility that there is more than one high affinity NGF receptor protein (Kaplan et aI 1 99 l ) . ECM Receptors

A number of potential receptors have been described for the ECM proteins, LN, FN, and the collagens (Edgar 1989). Of these, the receptors whose function is best correlated with the induction of neurite growth belong to a superfamily of cell-surface heterodimers known as the integrins (Hynes 1 987; Ruoslahti & Pierschbacher 1 987). Integrins are segregated into families based on different IX subunits that have a common /3 subunit. Antibodies to the /31 integrin subunit have been shown to inhibit neurite outgrowth on LN, FN, collagens, and complex ECMs (Bozyczko & Horwitz 1 986; Tomaselli et al 1 986; Hall et al 1987). In studies of cell adhesion to these proteins, it has been found that several different integrin heterodimers can recognize the same ECM protein, including receptors with distinct /3 subunits (e.g. Wayner & Carter 1 987; Sonnenberg et al 1 990; Tomaselli et al 1 990). In general, it is not known which particular integrins are involved in neurite growth of a neuronal cell type on a particular ECM protein (but see Tomaselli et al 1 990), however, it seems reasonable to expect that at least some of the transduction mechanisms discovered may be shared by different integrins. INTEGRINS

CAM Receptors

Members of the Ig superfamily often bind to other Ig superfamily molecules (Williams & Barclay 1 988), either to a second member protein (heterophilic interaction), or to themselves (homophilic interaction). Intercellular homophilic interactions have been demonstrated for most of the Ig-related CAMs listed in Table 1 (Edelman 1 986; Grumet & Edelman 1 988; Filbin et al 1 988). Homophilic interactions are also implicated in cadherin function (Takeichi 1 990; Nose et al 1 990). These observations suggest that for Ig-related CAMs and cadherins, the receptors for these molecules are the cognate molecules themselves in the neuronal membrane. The importance of such homophilic interactions in the inducHOMOPHILICRECEPTORS


1 29


tion of neurite outgrowth has been directly demonstrated in the case of L 1 and NCAM (Lemmon et al1989; Doherty et aI1990). HETEROPHILIC RECEPTORS For molecules like TAG- I, which are not trans membrane proteins, the receptors would be expected either to be other (heterophilic) molecules, or to be functionally coupled to a membrane protein, so that the signal they provide could be transduced into the activation of growth cone advancement. Recent studies suggest that while TAG- l can act as a homophilic adhesion molecule when transfected into Drosophila cells, its ability to promote neurite outgrowth is heterophilic. DRG cells are not inhibited from extending neurites on a substrate of TAG- l by treatment with phospholipase C, which cleaves TAG- l from their membranes (T. JesseJI, personal communication). Since TAG- l has an RGD sequence in its FN-type repeats, and since TAG- I is secreted (Karagogeos et al 1 99 1 ), perhaps the receptor for this outgrowth-pro moting activity is an integrin (currently under investigation, J. Dodd & T. JesseU, personal communication). There is a precedent for receptor interaction between integrins and Ig superfamily molecules in the inter action oflCAM - l with LFA- l (Springer 1 990).

Perspectives

As noted above, ECM proteins and CAMs seem to act synergistically in the induction of neurite growth. This suggests that different growth promoting receptors may transduce signals through common intracellular mechanisms (Bixby et al 1988). Naturally, signals promoting neurite growth must eventually converge on a common pathway to cause neurite elongation. On the other hand, the different classes of receptors discussed in this section have very different structures and seem to have distinct cellular localizations and functions (e.g. Geiger et al 1 985). In fact, recent evidence suggests that differences in transduction mechanisms are likely, both in the case of neurotrophic factors (Damon et a1 1 990) and substrate acting proteins (Bixby & Jhabvala 1 990). In sum, both common and unique mechanisms are likely to be employed, and these can only be sorted out by the examination of receptor interactions with a case-by-case approach.

SECOND MESSENGERS IMPLICATED IN NEURITE GROWTH It is safe to say that for the growth-promoting proteins thus far discussed the transduction mechanisms themselves are poorly defined. Because of the sparsity of information available, our discussion is limited to a brief examination of the intracellular signaling mechanisms that have been

130

BIXBY & HARRIS

implicated in the control of neurite growth, with reference to pertinent receptor interactions.


Calcium

It has been known for many years that manipulations affecting levels of intracellular Ca2-t affect neurite growth and growth cone motility, but the precise relationship between intracellular Ca 2 + levels and neurite growth has proven difficult to define. The data are consistent with the hypothesis that elevations of Ca 2 + are both stimulatory to neurite growth (Schubert et al 1978; Anglister et al 1982; Suarez-Isla et al 1984) and inhibitory (Bixby & Spitzer 1984; Mattson & Kater 1987). A theoretical resolution of this problem was proposed by Kater and colleagues, who suggested that an optimal Ca 2 + level exists in growth cones, and that raising or lowering Ca 2+ from this level will be inhibitory to growth (Mattson & Kater 1987). This hypothesis is attractive, but some data do not fit well (Silver et al 1989). A further complication is that different levels of Ca 2 + are achieved in different intracellular locations (i.e. neurite vs growth cone, different regions of the growth cone), and there may be unique Ca 2 + optima in different regions (Silver et al 1 990). Because Ca 2 + is a regulator of numerous important intracellular func tions, it would be surprising if Ca 2+ levels had no influence on neurite growth. Indeed, signals such as electrical impulses and neurotransmitters can elevate intracellular Ca 2 + and thereby affect neurite growth (e.g. Mattson & Kater 1987; Lankford et al 1988; Silver et al 1990). What is less clear is whether or not the control of intracellular Ca 2 + is an element of signal transduction used by growth-promoting proteins. One study using antibodies to CAMs to trigger transduction events found that intracellular Ca 2+ could be raised by incubating PCI2 cells with antibodies to L l or NCAM (Schuch et al 1989). Further studies will be necessary to allow interpretation of this result and to test its generality. cAMP

Like Ca 2 + , cAMP is a ubiquitous intracellular regulatory molecule that has been implicated in the control of neurite growth. As is also the case for Ca 2+ , studies linking intracellular cAMP levels to regulation of neurite outgrowth have led to differing conclusions. In a number of neuronal types responding to a variety of cues, elevations of intracellular cAMP have been shown to be inhibitory to neurite growth and/or growth cone motility (Forscher et al 1987; Lankford et al 1988; Mattson et al 1988; Bixby 1989). Experiments with neural cell lines, in apparent contrast, suggest that increasing cAMP levels is stimulatory for neurite growth (Nirenberg et al 1983; Greene et al 1986). However, this apparent discrepancy may be



131

due to a lack of distinction between neuronal axon-like neurites and the processes that are extended from tumor cells (Greene & Tischler 1 982; Greene et al 1 986). Whether cAMP is a second messenger for the transduction of the growth promoting signals under discussion has not been examined in most cases, but it does not seem likely at present. Elevation of cAMP is inhibitory to LN-stimulated neurite growth from parasympathetic neurons (Bixby 1 989), and a drug that inhibits cAMP-dependent protein kinase has no effect on this LN-induced growth (Bixby & Ihabvala 1 990). Antibodies to CAMs that do trigger other intracellular responses do not affect cAMP levels in PC1 2 cells (Schuch et al 1 989). Damon et al ( 1990) found that NGF and FGF can each promote neurite growth from PCl 2 cells that are deficient in cAMP responsiveness. G Proteins and

ras

Two different classes of GTP-binding proteins have been implicated in the regulation of neurite growth, although their relationship to transduction through growth-promoting proteins is not clear. The first of these is the oncogene, p21 v-ras. In PC1 2 cells, introduction of the v-ras protein promotes neurite extension (Bar-Sagi & Feramisco 1 985; Noda et aI 1 985), and injection of antibodies to p21 ras can block NGF-induced neurite growth (Hagag et al 1 986) . These results are consistent with a role for p 21 ras in the induction of neurites by NGF, at least in PCl 2 cells. More recently, Borasio et al ( 1 989) extended these results by showing that intra cellular application of the p2 1 ras-transforming protein (but not its normal cellular homologue) can potentiate neurite growth in genuine neurons. The fact that transformed cells and/or the transforming version of ras are apparently required to implicate ras in neurite growth is worth noting, as this makes the physiological relevance less obvious. However, these results may be merely a reflection of the exaggerated activity of an oncogene, which could be revealing an effect too subtle to be seen with the normal proto-oncogene in these relatively crude experiments. Although no evi dence exists that CAMs and ECM proteins signal neurite growth through ras, at least one member of the Ig superfamily may use ras as a messenger. Recently, Downward et al (1990) showed that activation ofT cells through the T cell receptor leads to activation of c-ras through the inhibition of GTPase-activating protein (GAP). The so-called large G proteins appear to be involved in the transduction of numerous transmembrane signals (see Casey & Gilman 1 988). The predominate G protein in brain is Go (Sternweis & Robishaw 1 984), and Go is concentrated near cell-cell junctions in vivo and at the tips of growing neurites in vitro (Strittmatter et al 1 990). The function of Go is not clear,


1 32

BIXBY & HARRIS

but it has been implicated, among other things, in the control of phos pholipase C. Two lines of indirect evidence suggest that Go may be involved in signal transduction leading to axon growth. First, pertussis toxin, which inhibits the function of Go and G;, blocks some of the responses to anti CAM antibodies in PC 1 2 cells (Schuch et al 1 989). Secondly, the axonal growth-associated protein GAP-43 (Skene 1989) has been implicated in the regulation of axonal growth, as seen in a large body of indirect evidence (for review, see Skene 1 989). It was recently shown that GAP-43 can stimulate the binding of GTPyS to Go (Strittmatter et al 1 990), which suggests an activating function. However, the relationship of this obser vation to the control of axon growth is not known. Tyrosine Kinases and

src

A number of growth factor receptors are protein tyrosine kinases, and other receptors have been shown to associate with tyrosine kinases that are themselves not receptors (e.g. Kypta et al 1990). Therefore, although there is no direct evidence that receptors for neuronal growth-promoting proteins are tyrosine kinases, the possibility exists that these receptors associate with the non-receptor tyrosine kinases of the src family (Cooper 1 989). The first evidence of this came from the observation that the v-src transforming protein can induce neurites in PC1 2 cells (Alema et aI 1 985). Subsequently, Maher ( 1 988) showed that NGF induces tyrosine phos phorylation in PC1 2 cells. These observations are consistent with the view that NGF may operate to produce neurites through a tyrosine kinase, at least in part. For example, the accessory protein required for the functional NGF-R could be a non-receptor type kinase. A second possibility is that the receptor tyrosine kinase trk comprises another high affinity NGF receptor (Kaplan et al 1 99 1 ). However, it may be that while tyrosine kinases are involved in the regulation of axon growth, they do not play a direct transducing role. Tyrosine phosphorylation of the /3 1 integrin recep tor subunit (in non-neuronal cells) leads to altered association with cyto skeletal elements (Tapley et al 1989). In Drosophila, the abl tyrosine kinase gene can be deleted without disrupting axonal growth, but if abl and another gene (dab) are both deleted, axonal guidance in the CNS is severely disrupted (Gertler et al 1 989; see also Elkins et aI 1 990). This evidence is in line with a modulatory role for tyrosine kinases in axonal growth. The actin-binding protein, vinculin, is a major substrate of tyrosine kinase activity in the growth cone (Igarashi et al 1990), which suggests a role in cytoskeletal organization. Recently, Mallen et al ( 1990) showed that tubulin is phosphorylated on tyrosine in neuronal growth cones (prob ably by c-src) and that this phosphorylation inhibits tubulin poly merization (W. Matten, P. Maness, personal communication). Since tubu-


1 33

lin polymerization is obviously critical for axon growth, phosphorylation by c-src may locally inhibit growth, perhaps by playing a role in axon guidance. Preliminary evidence using inhibitors of tyrosine phos phorylation to affect substrate-induced neurite growth in vitro is consistent with this possibility (J. Bixby, unpublished observations). In summary, the scanty evidence available suggests that tyrosine kinases are not the transducing elements for the growth signal, but probably do play a regu latory role.


Protein Kinase C

Protein kinase C (PKC) is probably the messenger most often linked to neurite growth-promoting signal transduction. PKC is in reality a family of proteins with at least seven isoforms, one or more of which are expressed predominantly in brain. PKC most likely participates in a variety of transduction events, usually as a consequence of the receptor-mediated activation of phospholipase C, which leads to the formation of the PKC activator diacylglycerol (for reviews of PKC, see Nishizuka 1 988; Huang 1 989). There are numerous papers documenting a relationship between PKC and the action of NGF. NGF activates PKC in responsive cells, and at least some inhibitors of PKC inhibit NGF-induced neurite growth (Traynor & Schubert 1 984; Hama et a1 1 986; Chan et al 1 989; Hashimoto 1 988; Koizumi et al 1 988). However, experiments using persistent appli cation of the phorbol ester TPA (tetradecanoyl phorbol l 3-acetate) to down-regulate PKC in PC 1 2 cells suggest that PKC function is not required for the neurite-inducing response to NGF or to FGF (Damon et al 1 990). PKC was not directly measured in these experiments, so this interpretation is not certain. At the moment, the weight of evidence sug gests that PKC is involved in transducing the neurite-promoting signal, but other pathways seem likely to be involved as well. It is natural to wonder whether PKC may also be involved in the response to ECM proteins and CAMs. This question becomes more rele vant because other members of the families involved (i.e. integrins and Jg superfamily molecules) may well employ PKC as a signal (Banga et al 1 986; Burn et a1 1 988; Cambier & Ransom 1 987; Desai et aI1 990). Experi ments using PKC inhibitors to test its involvement in substrate-induced neurite growth suggest the following conclusions. ECM proteins like LN, FN, and collagens induce neurites through a process dependent on PKC function, while CAM s (N-cadherin and LJ) may use both PKC-inde pendent (early) and PKC-dependent (late) pathways (Bixby 1 989; Bixby & lhabvala 1 990). Activation of PKC by the substrate was not measured in these experiments, therefore, whether PKC is actually transducing the signals remains an open question. Interpretation is further complicated by

1 34

BIXBY & HARRIS

the suggestion that integrin receptor function can itself be regulated by PKC (e.g. Shaw et al 1990; but see Danilov & Juliano 1989). If PKC does transduce neurite growth-promoting signals, what are the relevant substrates for phosphorylation? No direct evidence exists on this point. Two proteins associated with the polymerization of microtubules, MAP2 and tau, are PKC substrates, however, as is the growth-associated protein GAP-43 (Huang 1 989). The relationship of this observation to the finding that GAP-43 may activate Go is unclear. Annu. Rev. Cell. Biol. 1991.7:117-159. Downloaded from www.annualreviews.org by Pennsylvania State University on 08/02/12. For personal use only.

Nuclear Events

Substrate-acting neurite-promoting factors like LN and CAMs can induce neurites very rapidly. Nevertheless, these proteins can affect gene expression (e.g. Werb et al 1989), and the possibility of long-term effects on neurite growth remains to be addressed. For the neurotrophic factors, long-term effects on gene regulation have been known for years, and there is evidence that both transcription-independent and transcription dependent NGF pathways can induce neurites in PC12 cells (Greene & Shooter 1 980). The initial messengers involved in the control of gene expression by growth factors are not known. However, NGF and other growth factors can clearly induce the so-called immediate early genes such as c-fos, c-jun and jun- B (Wu et al 1 989; Bartel et al 1989). Of these, the c-jun induction shows at least some specificity (Bartel et al 1989). The challenge for the next few years will be to discover the mechanism of this induction and the targets of these early response gene products. Perspectives

Our knowledge of the transduction elements associated with neurite growth is clearly meager. Even for the neurotrophic agents, the signaling of which has been studied for some time, it is not possible to construct a pathway leading from receptor binding to control of the cytoskeletal proteins required for process extension. Nevertheless, there is reasonable evidence that the proteins listed in Table 1 transduce signals through mechanisms common to other, better understood receptor interactions. From what we know now, it seems profitable to focus on GTP-binding proteins, tyrosine kinases, and PKC for the present.

MOLECULES THAT GUIDE AXONS Guidance vs Growth Promotion

A factor that guides as opposed to one that simply promotes neurite outgrowth can be defined as an activity that is actually responsible in vivo for steering growth cones along their proper pathways. Steering may take



1 35

the form of selective fasciculation of axons or adhesion between growth cones and labeled pathways of guiding cells or ECM factors. It may also take the form of directional orientation up gradients of diffusible chemotropic agents or substrate-bound positional cues. Growth promoters or inhibitors, diffusible or bound, in the right spatio-temporal pattern and context can be guidance molecules. Guidance factors may also be molecules that have no neurite-promoting or inhibiting activity, but that cause axons on a growth-promoting substrate to turn. Guidance factors cannot properly by defined in standard dissociated cell culture conditions, but the possibility of guidance activity can be strongly suggested in complex culture situations where axons make a predictable or physiologically relevant choice while growing on an otherwise permissive substrate. The selective contacts made by individual filopodia and perhaps the tension generated by them seem to be important for growth cone guidance (Wessels & Nuttall 1978; Bray & Chapman 1985; Bray 1989; Heidemann et al 1 990). Growth cones in vitro may extend without filopodia, for instance in the presence of cytochalasin (Letourneau et al 1987), but in situ growth cones become disoriented, although they continue to extend, when filopodia are collapsed (Bentley & Toroian -Raymond 1 986). Another mechanism of growth cone turning may involve the selective retraction or micropruning of growth cone branches formed from lamellipodia (Bur meister & Goldberg 1 988). Receptor-mediated transduction as well as adhesion in the growth cone can be driven by many different molecular pathways, which raises the possibility that the growth cone may use sep arate pathways for guidance and growth. For instance, one might imagine that one pathway serves to promote growth by encouraging microtubule insertion into the base of the growth cone (Bamburg et al 1986), while another plays a more important role in guidance by causing asymmetrical cytoskeletal rearrangements at the leading edge of the growth cone. Similar suggestions could be made for actin-based motility (Smith 1 988). Recently, dynamic changes in the distribution of fluorescently labeled microtubules have been studied by J. Sabry et al (personal communication). They have seen that bundling of micro tubules on one side of the growth cone and the invasion of such bundles into particular growth cone branches can predict the direction of turning. It has been argued that signals received on one side of a growth cone can be amplified by a transduction cascade to create a local concentration gradient of an intermediate across the growth cone that is ultimately responsible for this asymmetric extension (Gierer 1 987; Walter et al 1990). As noted above, src could conceivably play a role in such a process. The factors that have been implicated in guiding axons are listed in Table 2. Because it is difficult to establish unequivocally a guidance func-

1 36

BIXBY & HARRIS

Table 2

Putative guidance factors

Factors

References

Cell surface Ig superfamily NCAM

Silver & Rutishauser ( 1 984) Fraser et a1 ( 1 988)


Landmesser et al ( 1 988, 1 990) Olfactory NCAM

Key & Akeson ( 1 990)

Fasciclin II

Harrelson & Goodman ( 1 988)

Ll

Dodd & Jessell ( 1 990)

Neuroglian

Bieber et al ( 1 989)

TAG- l

Dodd & Jessell ( 1 990)

Neurofascin

Rathjen et al ( 1 987)

Contactin

Ranscht ( 1 988)

Cadherins N-cadherin

Takeichi ( 1 990)

T-cadherin

Ranscht & Dours-Zimmerman ( 1 99 1 )

Others a l ( 1 990)

Fasciclin I

Elkins

Fasciclin III

Patel et al ( 1 987)

Defasciculation factor

Zipser et al ( 1 989)

et

ECM Laminin

Hammarback & Letourneau ( 1 986)

Tenascin/l l

Steindler et al ( 1 990) Snow et al ( 1 990)

Proteoglycans and glycoconjugates

Lander ( 1 989) Dodd & Jessell ( 1 987)

Diffusible NGF

Gunderson & Barrett ( 1 990)

Max factor

Davies & Lumsden ( 1 990)

Pons factor

HeITner et al ( 1 990)

Edwards et al ( 1 989)

Floor plate factor

Tessier-Lavigne et al ( 1 988) Placzek et al ( 1 990)

Inhibitory Central

Kapfhammer & Raper ( 1 987b)

Peripheral

Kapfhammer & Raper ( 1 987b)

Nasal

Raper & Grunewald (1 990)

Posterior sclerotome

Davies et al ( 1 990)

Posterior tectal

Stahl et al ( 1 990) Walter et al ( 1 990)

NI-35 and NI-250

Schwab ( 1990)

Serotonin and dopamine

McCobb ( 1 990)


Table 2

(continued)

Factors Topographic Posterior tecta I TRAP TOPDv


1 37

TOP A? "68K-Laminin Receptor" JONES ROCA-\

References

Walter et al (1 987) McLoon ( 1 99 1 ) Trisler et a l ( 1 98 1) Trisler & Collins (1 987) Trisler (1990) Rabacchi et al (\ 990) Constantine-Paton et al ( 1 986) Suzue et a\ ( \990)

tion for any factor, and because we believe some factors have both guidance and growth-promoting functions, we have included many of the factors from Table I in Table 2. We used a few simple criteria for assigning a factor a possible role in guidance. These criteria are reflected in the following questions. (a) Is the spatio-temporal expression of the activity in vivo strongly suggestive of a guidance function? (b) Does the block or (c) misexpression of the putative guidance molecule cause predictable mis routing, especially if blocking it does not inhibit further neurite extension? (d) In culture, is the patterned or localized application of the factor able to cause growth cones growing on a permissive substrate to change their direction? (e) Can this response be blocked with specific reagents? An answer in the affirmative t o one or more of these questions was sufficient for inclusion in Table 2 as a putative guidance factor. Of course, a negative answer to any of these questions does not rule out a guidance function. In most of the cases . listed, a definitive role in guidance has not been established. Putative guidance factors come in several different categories. First there are the growth promoters, which are cell surface bound, extraceIlular matrix bound, or diffusible. Second are a class of factors that are inhibitory for axon growth in vitro; i.e. factors that cause axon collapse or growth cone paralysis. These are obviously not growth promoting. Third, there are a few factors that should be considered as putative guidance molecules solely because of their topographic distribution, which matches the topo graphic distribution of certain sets of axonal projections. Finally, there is a set of nonchemical factors that has been proposed to have guidance activity. Chief among these are natural electric fields, and the strategic

placement or development of physical or morphological barriers. We refer readers to other reviews for these topics, as they will not be discussed

1 38

BIXBY & HARRIS

further here (Lockerbie 1 98 7; Kater & Letourneau 1 985; Letourneau 1 982).

Cell Surface Proteins NCAM was originally shown to have a role in axon fasciculation in vitro; a similar role in vivo might well argue for a guidance function of NCAM (Edelman 1 986). Evidence that NCAM does have a role in axon guidance in vivo comes from experiments in which functionally effective NCAM antibodies were injected into different regions of the developing and regenerating visual and motor system. Injections into the eye caused aberrations of optic fiber growth, particularly at the optic nerve head (Silver & Rutishauser 1 984), applications to the tectum caused optic axons to avoid the area of high antibody concentration (Fraser et al 1 988), and injections into the egg led to decreased branching in particular leg muscles in the chick (Landmesser et aI 1 988). The early, sustained, and widespread spatial distribution of NCAM in the CNS in vivo does not give much of an indication that it can act as a guidance molecule, and it may be argued that the in vivo effect of antibodies was nonspecific in terms of axon guidance because the entire tissue exposed to the antibody was compro mised. Silver et al ( 1 987), however, suggest that optic axons may avoid entering the telencephalon by encountering and being blocked by an NCAM-free "glial knot." The knot may also present physical barriers to axon growth so the relevance of NCAM to steering in this case is ques tionable. Furthermore, such NCAM-free zones do not seem to be common in the CNS. Since NCAM can be found in various states of sialylation, which changes developmentally in the brain, and which shows differential adhesivity (Edelman 1 986), it has been proposed that the changing state of this molecule rather than the distribution of the protein core can guide axons. Indeed, reductions in the sialylation state of NCAM can be correlated with innervation of target structures (Landmesser et a1 1 990; Schlosshauer et al 1 984). This argues that NCAM may be influencing target choice or branching pattern, for example, by making the target particularly adhesive and unsupportive of further growth; in a sense gluing the innervating axons into place. In line with this idea, enzymatic removal of the polysialic acid of NCAM led to increased fasciculation and decreased branching of ingrowing motor nerve terminals onto their target muscles (Landmesser et al 1 990). In addition, while reducing the sialylation state of NCAM increases cell adhesion (Rutishauser et al 1 98 8), fully unsialylated NCAM has almost no neurite outgrowth-promoting activity in culture (Doherty et aI 1 990). Key & Akeson ( l 990) recently reported that there is a uniquely glyco-


NCAM


1 39

sylated form of NCAM , recognized by a panel of monoclonal antibodies, that is expressed solely on the olfactory pathway. This highlights the question of whether or not there is a pattern of particular forms of the molecule present that correlates with and may guide specific axons, but at present reagents are too crude to resolve this matter. Several other members of the Ig superfamily such as L l (Chang et aI 1 987), F l l /contactin (Brum mendorf et a1 1 989; Ranscht 1 988), and TAG-l (Furley et aI 1990), as well as fasciclin II (Harrelson & Goodman 1988) and neuroglian (Bieber et al 1 989) in the fly, have a more limited distribution on specific axon bundles and thus may be more likely to have guidance functions. Few have been explicitly tested for guidance activity. In the mammalian embryo, TAG- l is expressed on the lateral edges of the spinal cord on the ventral-going segments of the axons of commissural descending interneurons, while L l is expressed on the posterior-going or descending segments of these neurons (Dodd & Jessell I988). The regional expression of different surface proteins along the length of a single axon (Dodd et al 1 988), when the transition from one factor to another occurs at a point on the axon where the trajectory changes dramatically (Bovolenta & Dodd 1 990), strongly sug gests that the change in expression of these molecules has a causal con nection with the change in direction, especially when these proteins can be demonstrated in vitro to have growth-promoting activity. The challenge now is to demonstrate a guidance function for TAG- l or L l in an in vivo or near in vivo situation.


L 1 , TAG- I , F I I ICONTACTIN, FASCICLIN II, NEUROGLIAN

The same considerations have led Goodman and colleagues to be particularly interested in axonal surface antigens that are expressed along part of an axonal trajectory of embryonic grasshopper neurons (Goodman et al 1 986). Such criteria helped focus attention on the fascic1ins. In the case of Drosophila fascic1ins, powerful expcriments are possible because of mutational analysis. It has, nevertheless, been difficult to demonstrate the in vivo significance of these molecules by mutational deletion analyses. The las I story is furthest advanced, with the las I mutants have a normal looking CNS at all stages examined, which indi cates that the molecule is not essential for either growth or guidance of any axons of the CNS (Elkins et al 1 990). Of course, this does not mean that las I is not a guidance molecule, but it does suggest that las I might be part of a redundant guidance system. To test this possibility, a series of double mutants was made with las I and several other mutants in putative guidance or growth cone function, each of which also had no phenotype on its own. Only one of the several tested showed neural defects in combination with fas I. This was the las I;abl double mutant. abl

FASCICLlN I


1 40

BIXBY & HARRIS

codes for a tyrosine kinase (see above), which may function in a signal transduction pathway in the growth cone. That both genes had to be knocked out to achieve axon disorientation suggests that the two genes are part of two separate pathways, either of which may suffice for guidance (Elkins et al 1 990). In the grasshopper limb, Jay & Keshishian ( 1 990) recently were able to knock out Jas I in vivo by a new and potentially powerful technique called chromophore-assisted laser inactivation (CALI). In contrast to the mutational analysis, however, they found that the inactivation of Jas I alone disturbed axon guidance. It caused pioneer axons to wander and defasciculate, although in the end they were able to find their way into the CNS. Why could CALI reveal an effect where mutational analysis did not? First, CALI might inactivate entire molecular complexes while mutations knock out one molecule at a time. Second, mutants might have time to compensate for the loss ofJas I, whereas the CALI knock-out immediately preceded axon growth. Finally, there is the possibility that thejas I mutants might have shown an effect in the fly leg disc that was not examined, as the CNS of the CALI-treated grasshopper was also not examined. While these details have to be sorted out, it is clear from both of these studies that Jas I meets sufficient conditions to call it a true guidance molecule in vivo. Indeed, it may be the first to achieve such status. The ultimate experiment, of expressing Jas I in inappropriate tracts, for instance, by putting it under the control of a different fasciclin promoter has yet to be done. Predictable misdirecting of particular axons should occur in this situation. FASCICLIN I I AND III The proteins fasciclin II and fasciclin III were isolated and cloned from Drosophila on the basis of a characteristic distribution on subsets of developing axons (Patel et al 1 987; Harrelson & Goodman 1 989; Snow et al 1 989; Elkins et al 1 990). DNA transfection studies have shown that fasciclin III, like fasciclin I, can function as a homophilic cell adhesion molecule, similar to members of the Ig superfamily or the cadherins (Elkins et aI 1 990). The localization and CAM structure (fas II) or function (fas III) strongly suggest that they too will be involved with axonal growth and guidance. CADHERINS The cadherins form another set of putative neuronal guidance proteins. N-cadherin, which in early stages of development seems widely distributed throughout the nervous system, becomes restricted to certain pathways such as the retina and optic tract during the differentiation of axonal tracts (Takeichi 1 988). This restriction is not really sharp enough to suggest that N-cadherin is a guidance molecule. There is another neural cadherin, however, called T-cadherin, which is expressed in a very restric-


141


ted pattern, i.e. o n motor neurons, o n the floor plate, and o n the posterior sclerotomcs of chick embryos (Ranscht & Dours 1 989; Ranscht & Bronner Fraser 1 991). This distribution makes it a candidate for a guidance molecule, perhaps inhibitory in nature (see below). Finally, there is R cadherin, which is expressed solely on the vitreal surface of the retina, and may promote or guide the intraretinal growth of optic fibers (M . Takeichi, personal communication). Whereas the function of N-cadherin in axon growth has been established, we emphasize that there is no evidence to support the notion that T- and R-cadherin play a role in this process, other than this patterned expression. While many of the cell surface molecules discussed above can be thought of as fasciculation-inducing molecules, Zipser et al (1 989) reported a carbohydrate determinant that is expressed on 1 30-kd glycoproteins in sensory afferents of leech neurons, and which appears to be involved in the defasciculation of these fibers as they enter the eNS. A monoclonal agent against this determinant prevented the defasciculation in organ culture, and the sensory afferents failed to expand their axonal arbors when they entered the CNS. Whether this is a dedicated defasciculation factor or a molecule like NCAM, with the adhesive state modified by the carbohydrate moiety, is unclear.

LEECH DEFASCICULATION FACTOR

ECM Molecules In vitro LN clearly has promoting activity (see above). When applied as a patterned substrate, in vitro growth cones can follow LN trails, preferring LN to other growth-promoting substrates, such as FN, when given a choice (Vielmetter et al 1 990). Axons can skip from one region of LN to another over areas of low adhesivity (Hammarback & Letourneau 1 986). These culture results show that it is indeed feasible for LN to act as a guidance molecule. Concentration gradients of LN have also been tested in vitro to see if it is possible that an adhesive gradient might guide axons. Even though the gradients studied were steep and extended through the physiological range of activity, they did not seem to be capable of steering growth cones (McKenna & Raper 1 988). These results make it likely that if LN is acting as a guidance molecule in vivo, it is doing so by its presence or absence, rather than by a graded distribution. Interestingly, however, the unc-6 mutant of C. elegans has a defect in dorso-ventral guidance and is mutated in the B2 chain of laminin, which suggests that laminin is somehow responsible for topographic guidance in this system (reviewed in Hedgecock & Hall 1 990). The problem in the vertebrate is that LN seems to be all over the basal lamina and the endfeet of glial cells in the eNS and thus does not form any patterns reminiscent of particular LN

1 42

BIXBY & HARRIS


pathways, except that the pial surface of the brain seems to be where most long projecting axons of the CNS grow during development. Moreover, LN appears to be an excellent substrate for most, if not all, central neurons (Lander 1 987). These two facts argue that LN is largely playing a per missive rather than an instructive role in axon navigation, at least in the CNS. However, until more is learned about the distribution and exposure of the different forms of laminin, the different binding sites on LN, and the expression of different integrin complexes and LN receptors (see above), we cannot be certain of this conclusion. F N , TENASCIN, PROTEOGLYCANS, GLYCOCONJUGATES FN is clearly a growth promoting ECM factor, but its rather general immunohistochemical local ization in the nervous system suggests that it too is not an important guidance molecule (Rogers et al 1 989), and experiments in vivo do not suggest that FN has guidance activity. Tenascin is another transiently expressed ECM molecule in the brain with a more restricted pattern, particularly around central glia (Snow et al 1 990). Like T-cadherin, it is restricted to half somites, but does not seem to influence the growth of motorneurons (Stern et aI 1 989). It has been argued that the distribution of tenascin surrounding the barrel columns in the developing somatosensory cortex serves a role in restricting afferents or otherwise limiting horizontal connections within individual barrels (Steindler et aI 1 990). In vitro studies oftenascin show it to be a complicated multifunctional molecule with both adhesive and anti-adhesive activities, depending perhaps on the functional domain exposed (Spring et a1 1 989; Faissner & Kruse 1 990). While growth cones respond to tenascin, there is no evidence to date that it can support or inhibit neurite outgrowth. Proteoglycans (PGs) are an extremely diverse group of molecules composed of glycosaminoglycan (GAG) chains and protein cores. These PGs, such as heparin sulfate PG, can by themselves promote neurite outgrowth (Hantaz-Ambroise et al 1 987), but they are often found in combination with ECM molecules (Lander 1 989). The potentially vast array of PG-ECM molecule complexes makes this an intriguing possibility for a combinatorial molecular system with enough complexity to guide a variety of axons along different pathways (Herndon & Lander 1990). There is also a large variety of glycoconjugates on the surfaces of subsets of neurons (Dodd & Jessell 1 986; Jessell et al 1 990), which may also have a role in axon growth and guidance. Before we take these ideas too seriously, it is critical to know much more about the distribution and activities of these complexes.

Diffusible Neurotropic Factors

The only identified protein demonstrated to be chemo tropic to axons is NGF. In vitro, NGF can keep the growth cone-tipped

NGF, MAX FACTOR



1 43

axons of sympathetic cells from retracting (Campenot 1 982a,b); moreover such growth cones, while growing on a permissive substrate, will turn up concentration gradients of NGF (Gunderson & Barrett 1 980). Gunderson & Barrett ( 1980) also demonstrated that growth cones would turn toward dB cAMP released from a nearby micropipette, which indicates a possible role of cAMP as an intermediate in NGF-directed guidance. Intracellular Ca 2 + release was also implicated in their experiments, since growth cones would turn up concentration gradients of Ca 2 + in the presence of the Ca 2 + -ionophore, A23 1 87, while dantrolene, which blocks the intracellular release of Ca 2 + , blocks turning toward either NGF or dB cAMP (Gunder son & Barrett 1 980). In vivo we know that NGF is manufactured and secreted from targets (Davies et al 1 987). Further, when NGF is injected into the brain, it leads to unusual central projections of sympathetic fibers toward the source of injected NGF (Menesini-Chen et al 1 978), and when it is expressed from a transgene in fibroblasts injected into the brain, it can lead to the sprouting of cholinergic neurons towards the transplanted cells (Rosenberg et al 1 988). Edwards et al ( 1 989) showed in transgenic mice that pancreatic beta cells expressing NGF from a heterologous promoter attract extra sympathetic innervation. In spite of this extremely powerful evidence, it is still uncertain whether NGF has a real tropic role in vivo. It can be argued that the levels of NGF needed to make axons turn in vitro and in vivo are not physiological. Moreover, NGF has more or less been ruled out to be a guidance molecule in the trigeminal sensory system, where sensory axons are chemotropically guided to the maxillary whisker pad of the epithelium (Davies & Lumsden 1 990). In this case it appears that the NGF receptor is up-regulated in the sensory neurons, but only after the neurons arrive at their target, consistent with a role for NGF as a trophic factor, shaping innervation by restricting neuronal survival rather than guiding axons (Wyatt et aI 1 990). This strongly suggests the existence of an as yet unidenti fied chemotropic factor (tentatively called Max factor) that is responsible for this attraction (Davies & Lumsden 1 990). Chemotropic studies have not yet been reported for the more recently discovered members of the NGF family of neurotrophic molecules. Recently, other instances of tropic guid ance have been demonstrated in the CNS. For example, cortical fibers are attracted to the pons (Heffner et al 1 990), and commissural interneurons of the spinal cord are attracted to the floor plate (Tessier-Lavigne et al 1 988; Placzek et al 1 990). That these are instances of chemotropism has been established in co-cultures imbedded in three-dimensional collagen gels. The floor plate factor appears to be a large protein, about 90 kd, that FLOOR PLATE AND PONS FACTORS

1 44

BIXBY & HARRIS

may possibly bind to the ECM (T. Jessell & M . Tessier-Lavigne, personal communication), characteristics which distinguish it from NGF. The actual diffusible factors released from the maxillary pad, the floor plate, and the pons have not yet been identified, and it will be intriguing to learn more about their biochemical characteristics. Are they, for instance, related to each other? At the moment, we can only say that the existence of such chemotropic guidance mechanisms in the nervous system seems certain.


Inhibitory Factors COLLAPSING FACTORS The notion that axons could be guided by inhibitory rather than growth-promoting factors was rekindled in our consciousness only a few years ago when Kapfhammer & Raper ( 1 987a) demonstrated that the growth cones of retinal cells collapse when they encounter the axons of sympathetic cells and vice-versa. By testing a variety of different types of axons against each other, Kapfhammer & Raper ( 1987b) demon strated that there are a minimum of three collapsing factors among the several classes of neurons tested: a central collapsing factor, a peripheral one, and one on nasal, but not temporal, retinal axons (Raper & Grune wald 1 990). The actual number of distinct collapsing factors in the brain may be considerably larger. Collapsing activity can be easily demonstrated in vitro by sprinkling membranous vesicles of test material onto growing axons (Raper & Kapfhammer 1 990). Using this assay, two collapsing activities, one from posterior sclerotome tissues (Davies et al 1 990) and one from posterior tectum, have been identified (Stahl et al 1 990), and another from brain has been greatly enriched (J. Raper, personal com munication). All these factors are different in terms of their molecular weights, lectin-binding capacity, and susceptibility to urea (Walter et al 1 990). Whether there are families of such factors will be revealed only after these and others are identified and sequenced. The reason for thinking that inhibitory factors are true guidance mol ecules has to do first with their distribution in vivo. For instance, both the posterior tectal factor and posterior sclerotome factors are expressed in a pattern that by itself can account for the absence of particular axons entering these regions. A first test to see whether these factors truly have guidance function in vivo will be to block the activity with antibodies. This should lead to the invasion of new areas by spinal axons and temporal retinal fibers. Ca 2 + has been measured using fura-2 as a possible intermediate in growth cone collapse (Ivins et al 1 990). However, Ca 2 + levels were rela tively constant through this dramatic change in growth cone morphology. T-CADHEt{IN, NI-35, NI-250

T-cadherin also may have an activity in repelling


145


spinal axons from entering the posterior somite. T-cadherin is GPI-linked (Ranscht & Dours-Zimmerman 1 99 1 ) and thus lacks an intracellular domain that is critical for the adhesive functions of cadherins (Nose et al 1 990). This may be the basis of its putative avoidance function. Schwab ( 1 990) identified two factors (NI-35 and NI-250) on oligodendrocytes that inhibit or paralyze growth cone advance. These factors have not been postulated to have a role in axonal guidance, and their normal function in vivo is still a mystery. It seems likely, however, that their activity is partially responsible for preventing the regeneration of axonal tracts in the mammalian eNS. Another class of growth-inhibiting factor is neuro transmitters, which when applied to the growth cones of specific growing Helisoma neurons in culture, paralyzes them (McCobb et aI 1 990), perhaps by opening Ca 2 + channels (see above). The growth cones paralyzed by serotonin come from neurons that normally make contact with seroton ergic postsynaptic cells and can, therefore, be expected to receive stop and synapse cues from these postsynaptic partners. Few would have guessed that the growth cones of the presynaptic cell would be sensitive to the synaptic transmitter of its postsynaptic partner, but these findings suggest the possibility that such a feedback system may indeed be used in the nervous system. Low concentrations of serotonin can admit Ca 2 + and block growth cone advance in organ culture (Murrain et aI 1 990). More over, serotonin is expressed in the developing nervous system at a time when it could participate in growth cone guidance, and 5,7-dihydroxy tryptamine treatment of the embryo leads to aberrations in the growth pattern of the specific neurons whose growth cones are stopped by sero tonin in vitro (Goldberg & Kater 1 989).

TRANSMITIERS

s-LN is an ECM protein localized selectively at motor end plates i n vivo. In vitro, it can promote cell adhesion, and its localization in vivo suggests that it too may be a cue to stop and synapse (Hunter et aI 1 989).

s·LN

Topographic Guidance Molecules POSTERIOR TECTAL FACTOR, TRAP A number of molecules are good can didates for guidance factors not because of their presence on certain fascicles or tracts i n the brain, but solely because of their patterned distribution. If a neural tissue appears histologically uniform and yet i t displays a gradient of a cell surface marker that in some way parallels the distribution of axonal projections, we must consider the possibility that the molecule is a guidance factor. We have already discussed one of these, the inhibitory factor of the posterior tectum. Interestingly, the choice of temporal retinal cells not to grow on posterior tectal membranes is not graded in the retina,


1 46

BIXBY & HARRIS

but sharp. At approximately the center of the eye, there is a clear boundary between nasal and temporal retinal axons, and it is only the temporal retinal axons that avoid the posterior tectal membranes (Walter et aI 1 990). Until the distribution of the avoidance factor is demonstrated in vivo as a gradient, we must consider the possibility that it too is discontinuously distributed, perhaps along a curved or slanted line that separates anterior and posterior tectum. This avoidance protein was found (Stahl et a1 1 990) using the growth cone collapsing bioassay of Raper & Kapfhammer ( 1 990). More often, however, topographically distributed molecules in the nervous system have been found not by a functional assay, but by monoclonal antibodies. Along with the functional dichotomy of the retina into nasal and temporal halves, McLoon ( 1 9 9 1 ) recently described a protein called TRAP, which is distributed almost in a step function across the chick retina and mainly expressed in the temporal retina. Trisler and colleagues (Trisler et al 1 98 1 ; Trisler 1 990) used selective immunization and screening of monoclonal antibodies to identify two antigens that are expressed as orthogonal gradients on cell surfaces and ganglion cell axons of the chick retina. The first, called TOPov, a 47kd protein, is 35-fold higher in the dorsal than ventral retina. The second, called TOPAP, a 40-kd protein, is 1 6-fold higher in the anterior (nasal) than posterior (temporal) retina. The concentration of these TOP molecules varies continuously and logarithmically across the retina (Trisler 1 990). Both of these antigens are present in the tectum as well as the retina (Trisler & Collins 1 987; Trisler 1 990). In the tectum they are present as gradients that are complementary to the retinal gradients. TOPov is to-fold higher in the ventral (lateral) tectum, and TOPAP is 8-fold higher in the anterior tectum. The retinal gradients persist through the development and matura tion of the retina, while the tectal gradients become visible just prior to retinal innervation (Trisler 1 990). The finding of reciprocally graded, orthogonally expressed molecules in the retina and tectum is a prediction of Sperry's ( 1963) hypothesis of cytochemical chemospecificity, partic ularly as modified by proponents of antagonistic gradients of adhesion molecules and their receptors (e.g. Fraser & Hunt 1 980; Bonhoeffer & Gierer 1 984). Whether or not these molecules actually serve the roles suggested by their distribution has yet to be tested, but functional studies of TOPov within the retina, obtained by injecting antibodies intraocularly, suggest that TOPov has an effect on growth cone persistence and synapto genesis (Trisler et al 1 986; Trisler 1 990). TOPDv, TOPAP

Rabacchi et al ( 1990) recently reported two antibodies that stain the embryonic retina and tectum in a strongly graded fashion, from dorsal to ventral in the retina and lateral to medial in the tectum.

"68K-LN RECEPTOR"



147

These antibodies recognize a molecule previously identified as a " 68K-LN receptor" (Wewer et al 1 987), but which, as indicated by the quotation marks, is likely to have other functions. Interestingly, the protein recog nized by the antibodies and its mRNA are equally distributed from dorsal to ventral. It is the conformation of the antigen that appears graded. While TOPDv with its apparent relative molecular mass of a 47 kd would seem biochemically distinct from the "68K-LN receptor," Western blots of the "68K-LN receptor" using antibodies that recognize the topographically graded antigen reveal a predominant band of about 44 kd, which suggests that the mRNA might be spliced or that the 68 K form of the protein is post-translationally processed. The similarity of the molecular weights of TOPDV and the graded "68K-LN receptor" thus leave open the question of whether the two antigens are related. JONES is the name given to another dorsal to ventral-graded antigen in the developing retina (Constantine-Paton et al 1 986). JONES is a ganglioside. Its temporal as well as its spatial pattern of expression on axons is again suggestive of a chemoaffinity molecule, but it may instead have a role in the sequential maturation of the retina. Again, functional studies are needed to define its biological role.

JONES

ROCA - l Suzue et al ( 1 990) recently described an antigen called ROCA- I , a 65-kd protein on the surfaces o f glial cells o f the sympathetic chain and on the Schwann cells of the intercostal spinal nerves of the rat. ROCA-I 's abundance is strongly graded from rostral to caudal along the length of the animal. It has long been known that there is a reproducible rostrocaudal topography of sympathetic inputs, which is reestablished after regen eration, suggestive of graded chemospecific ity. The expression pattern of ROCA- l engenders some hope that it is in fact a guidance factor that helps establish the topographically appropriate connections of the sympathetic chain. If axons are using positionally graded cues to navigate, perhaps the genes that code for molecules like ROCA- l are under the direct control of spatially orchestrated patterns of transcription factors, like the homeobox containing genes that seem to be responsible for segmentation or anterio posterior axis formation. If so, this would mean that the positional infor mation, which guides the fate of cells in the nervous system could also be used to produce topographic displays of a guidance factors (Patel et al 1 989). Axons could then navigate through the brain to distant targets using these topographic cues like a map. Axoils with different final des tinations could use the same map but, by dint of their own particular topographic origin (and hence receptor types), could be guided to different targets with different positional values (Harris 1 989).

1 48

BIXBY & HARRIS


Perspectives

Unlike the situation for growth-promoting factors discussed above, the number of proven guidance molecules is very small, so it is unclear how many families exist and impossible to hazard even a wild guess as to their total numbers. It is also unclear how much overlap there is between guidance and growth-promoting factors. We know even less ahout the transduction pathways that lead to growth cone steering. It may be based on a receptor-mediated transduction cascade or even a stretch-activated mechanism. The number or different categories of putative guidance fac tors suggest, however, that the molecules involved may be a more varied and complex group than are the growth promoters. Unfortunately, because the assays for guidance often require in vivo analysis, the going will be harder. One way investigators are trying to improve their ability to deal with this difficulty is to set up in vitro systems that preserve natural guidance. The establishment of manipulatable in vivo systems or in vitro systems where guidance is separable from growth is the basis for establishing guidance function. Experimentally, the situation is a difficult one. A simple culture system, with neurons growing on plastic or LN in a defined chemi cal medium is easy to manipulate experimentally, but it does not come close to simulating the way axons grow and normally respond in vivo. In addition, what guides axons in a dish of dissociated cells may not be relevant to axon growth or guidance in vivo. Moreover, real axonal guid ance might rely on complex spatial and temporal arrays of guidance factors that are difficult to recreate in vitro. The embryo obviously provides real guidance cues, but is not as accessible experimentally. The embryo provides barriers to visibility, antibody penetration, and chemical manipulation, although there was stunning work done on visualizing growing neurons in the amphibian tail half a century ago (Speidel 1 933). In some cases manipUlations would be lethal to the embryo and thus compromise the experiment. The arguments above are abstracted from Landmesser ( 1 988), who is advocating the use of culture systems that on the one hand try to conserve the in vivo environment of the growing axon, and on the other, open the system up for visualization and various forms of experimentation not possible in vivo. Of course, the more the culture system takes on aspects of the true in vivo situation, the more confident one is that the guidance cues one is investigating are physiological, but the more difficult the system is to characterize chemically. Nevertheless, these halfway sys tems like the cultured whole retina (Halfter & Deiss 1984), the developing live slices of the chick limb bud (Landmesser' 1 988), or thalamocortex (Bolz et al 1 990), as well as the filleted grasshopper limb bud in culture



1 49

(Lefcort & Bentley 1 987; Condic & Bentley 1 989; O'Connor et al 1 990), and the exposed amphibian brain (Harris et aI 1 987), have begun to provide us with manipulatable systems in which real guidance is demonstrable. Molecular and genetic technology provide other powerful ways of test ing growth and guidance factors in vivo. In flies and worms, various mutants have been found or constructed that seem to perturb axon growth, fasciculation or guidance (e.g. Hedgecock & Hall 1 990; Boschert et al 1 990). Such genetic approaches may define new components of guidance. They may also provide' a strategy for testing already identified factors in vivo, as was done with fasciclin I (Elkins et al 1 990). Homologous recombination techniques with embryonic stem cel1s may make the latter approach feasible in the mouse. In many species it may be possible to express a genetically engineered mutant form of the putative guidance molecule transgenetically for in vivo analysis. While the latter technique has been successfully applied to study other cellular interactions, particularly developmental ones in mice and frogs, such studies are only beginning with a view to axon growth and guidance in vivo (Holt et al 1 990; Edwards et al 1 990). CONCLUSIONS

In the last ten years, we have learned a great deal about various molecular factors that play roles in neurite outgrowth and guidance. We have identi fied several families of molecules that, promote neurite outgrowth . We have made inroads into the transduction cascades of the growth cone, identifying receptors, second messenger systems, and motile cytoskeletal elements. We have also begun to classify the types of guidance cues that exist in the embryonic nervous system. What can we expect in the next ten years? Perhaps we shall come close to completing a list of the general growth-promoting molecules in at least the best-studied systems. We should find the missing links in the chain of sensory-motor activation in the growth cone, at least to a first approximation. And we shall, using in vivo or near in vivo systems, like the ones described above, identify a great many more authentic guidance factors and begin to understand the logic of their deployment in the embryonic nervous system. ACKNOWLEDGMENTS

This work was supported by National Institutes of Health awards to JLB and WAH, and a National Science Foundation award to JLB. We thank Christine Holt and Chi-Bin Chien for their helpful comments on the manuscript.

1 50

BIXBY

&

HARRIS


Literature Cited

Abraham, J. A., Mergia, A., Whang, J. L., Tumolo, A., Friedman, J., et al. 1986. Nucleotide sequence of a bovine clone encoding the angiogenic protein basic fibroblast growth factor. Science 235: 54548 Alema, S., Casal bore, P., Agostini, E., Tato, F. 1 985. Differentiation of PCI 2 phaeo chromocytoma cells induced by v-src oncogene. Nature 3 1 6: 557-59 Anglister, L., Farber, I. c., Shahar, A., Grin vald, A. 1982. Localization of voltage-sen sitive calcium channels along developing neurites: Their possible role in regulating neurite elongation. Dev. Bioi. 9 1 : 351-65 Bamburg, J, R" Bray, D" Chapman, K, 1986. M icrotubule assembly in the axon. Nature 323: 400 Banga, H . , Simons, E. R., Brass, L. F., Rit tenhouse, S. E. 1 986. Activation of phos pholipases A and C in human platelets exposed to epinephrine: Role of glyco proteins lIb/IlIa and dual role of epi nephrine. Proc . Natl. A cad. Sci. USA 83: 9 197-9201 Barde , Y. A. 1988. What, if anything, is a neurotrophic factor? Trends Neurosci. 11: 343-46 Bar-Sagi, D., Feramisco, J. R. 1985. Micro injection of the ras oncogene protein into PCI2 cells induces morphological differ entiation. Cell 42: 841-48 Bartel, D. P., Sheng, M., Lau, L. F., Green berg, M . E. 1989. Growth factors and membrane depolarization activate distinct programs of early response gene ex pression: Dissociation of fos and jun induction. Genes Dev. 3: 304- 1 3 Bastiani, M . J . , Doe, C . Q . , Helfand, S . L., Goodman, C. S. 1985. Neuronal speci ficity and growth cone guidance in grass hopper and Drosophila embryos. Trends Neurosci. 8: 257-66 Bentley, D., Toroian-Raymond, A. 1986. Disoriented pathfinding by pioneer neu rone growth cones deprived of filopodia by cytochalasin treatment. Nature 323: 712-15 Berg, D. K. 1984. New neuronal growth fac tors. Annu. Rev. Neurosci. 7: 149-70 Bieber, A. J., Snow, P. M . , Hortsch, M . , Patel, N . H . , Jacobs, J. R., e t al. 1989. Drosophila neuroglian: A member of the immunoglobulin superfamily with exten sive homology to the vertebrate neural cell adhesion molecule L l . Cell 59: 447-60 Bixby, J. L. 1 989. Protein kinase C is involved in laminin stimulation of neurite outgrowth. Neuron 3: 287-97 Bixby, J. L., Jhabvala, P. 1990. Extracellular matrix molecules and cell adhesion mol-

ecules induce neurites through different mechanisms. J. Cell Bioi. I I I : 2725-32 Bixby, 1. L., Lilien, J., Reichardt, L. F. 1988. Identification of the major proteins that promote neuronal process outgrowth on Schwann cells in vitro. J. Cell BioI. 1 07 : 353-62 Bixby, 1. L., Pratt, R. L., Lilien, J., Reich ardt, L. F. 1987. Neurite outgrowth on muscle cell surfaces involves extracellular matrix receptors as well as Ca 2 + -depen dent and independent cell adhesion mol ecules. Proc. Nat!. Acad. Sci. USA 84: 2555-69 Bixby, J. L., Spitzer, N. C. 1984. Early differentiation of vertebrate spinal neu rons in the absence of voltage-dependent Ca 2 + and Na+ influx. Dev. Bioi. 106: 89-96 Bixby, J. L., Zhang, R. 1 990. Purified N cadherin is a potent substrate for the rapid induction of neurite outgrowth, J. Cell BioI. 1 10: 1253-60 Boschert, U., Ramos, R., Tix, S., Technau, G., Fischbach, K.-F. 1 990. Genetic and developmental analysis of irreC, a genetic function required for optic chiasm for

mation in Drosophila. J. Neurogen. 6: 15371 Bolz, J., Novak, N., Gotz, M . , Bonhoeffer, F. 1990. Formation of target-specific neuronal projections in organotypic slice cultures from rat visual cortex. Nature 346: 349-62 Bonhoeffer, F., Gierer, A. 1984. How do retinal axons find their targets on the tec tum? Trends Neurosci. 7: 378-8 1 Borasio, G. D., John, J., Wittinghofer, A., Barde, Y.-A., Sendtner, M., Heumann, R. 1989. Ras p 2 1 protein promotes survival and fiber outgrowth of cultured embry onic neurons. Neuron 2: 1 087-96 Bovolenta, P., Dodd, J. 1990. Guidance of commissural growth cones at the floor plate in embryonic rat spinal cord. Development 1 09: 435-47 Bozyczko, D., Horwitz, A. F, 1 986. The par ticipation of a putative cell surface recep tor for laminin and fibronectin in peri pheral neurite extension. J. Neurosci. 6: 124 1-5 1 Bray, D. 1 989, Growth cone formation and navigation: Axonal growth. Curro Opin. Cell Bioi. I : 87-90 Bray, D., Chapman, K. 1985. Analysis of microspike movements on the neuronal growth cone. J. Neurosci. 5: 3204-13 Brummendorf, T., Wolff, J. M., Frank, R., Rathjen, F . G. 1 989. Neural cell recog nition molecule F I I : Homology with fibronectin type III and immunoglobulin type C domains. Neuron 2: 1 35 1-61



1 51

Burgess, W. H . , Maciag, T. 1 989. The hepa rin-binding (fibroblast) growth factor family of proteins. Annu. Rev. Biochern. 58: 575-606 Burmeister, D. W., Goldberg, D. J. 1 988. Micropruning: The mechanism of turning of Aplysia growth cones at substrate bor ders in vitro. J. Neurosci. 8: 3 1 5 1-59 Burn, P., Kupfer, A., Singer, S. J. 1 988. Dynamic membrane-cytoskeletal inter actions: specific association of integrin and talin in vivo after phorbol ester treat ment of peripheral blood lymphocytes. Proc. Natl. Acad. Sci. USA 85: 497-501 Cambier, J. C., Ransom, J. T. 1 987. Molec ular mechanisms of transmembrane sig nalling in B lymphocytes. Annu. Rev. lrnrnunol. 5: 1 75-200 Campenot, R. B. 1 982a. Development of sympathetic neurons in compartment alized cultures. I. Local control of neurite growth by nerve growth factor. Dev. BioI. 93: 1-12 Campenot, R. B. 1 982b. Development of sympathetic neurons in compartment alized cultures. I I . Local control of neurite survival by nerve growth factor. Dev. Bioi. 93: 1 3-2 1 Casey, P. J., Gilman, A. G. 1 988. G protein

tein-tyrosine kinases. I n Peptides and Pro tein Phosphorylation, ed. B. Kemp, pp. 86-1 1 3 . Boca Raton, Fla: CRC Press Crittenden, S. L., Rutishauser, U., Lilien, J. 1 988. Identification of two structural types of Ca + + -dependent cell adhesion mol ecules in the chick. Proc. Nat!. A cad. Sci. USA 85: 3464--68 Cunningham, B. A., Hemperiy, J. J., Murray, B . A., Prediger, E . A., Brack enbury, R., Edelman, G. M. 1 987. Neural cell adhesion molecule: structure, im munoglobulin-likc domains, cell surface modulation, and alternative RNA splic ing. Science 236: 799-803 Damon, D. H . , D'Amore, P. A., Wagner, J. A. 1 990. Nerve growth factor and fibro blast growth factor regulate neurite out growth and gene expression in PC 1 2 cells via both protein kinase Ca- and cAMP independent mechanisms. J. Cell BioI. 1 10: 1 333-39 Danilov, Y. N., Juliano, R. L. 1 989. Phorbol ester modulation of integrin-mediated cell adhesion: A post receptor event. J. Cell BioI. 1 08: 1 925-33 Davies, A. M . , BandtIaw, c., Heumann, R., Korsching, S., Rohrer, II., Thoenen, I I . 1 987. Timing and sites o f nerve growth

J. BioI. Chern. 263: 2577-80 Chan, B. L., Chao, M. V., Saltiel, A. R. 1 989. Nerve growth factor stimulates the hydrolysis of glycosyl-phosphatidyl inositol in PC- 1 2 cells: A mechanism of protein kinase C regulation. Proc. Natl. Acad. Sci. USA 86: 1 756--60 Chang, S., Rathjen, F. G., Raper, J. A. 1 987. Extension of neurites on axons is impaired by antibodies against specific neural cell surface glycoproteins. J. Cell BioI. 1 04: 355-62

relation to innervation and expression of the receptor. Nature 326: 353-58 Davies, J. A., Cook, G. M. W., Stern, C. D., Keynes, R. J. 1 990. Isolation from chick somites of a glycoprotein fraction that causes collapse of dorsal root ganglion growth cones. Neuron 4: 1 1-20 Davies, A. M., Lumsden, A. 1990. Ontogeny of the somatosensory system: Origins and early development of primary sensory neurons. Annu. Rev. Neurosci. 1 3 : 6 1-73 Desai, D. M . , Newton, M. E., Kadlecek, T., Weiss, A. 1 990. Stimulation of the phos phatidylinositol pathway can induce T cell activation. Nature 348: 66--69 Dodd, J., Jessell, T. M. 1 988. Axon guidance and the patterning of neuronal projections in vertebrates. Science 242: 692-99 Dodd, J., Jessell, T. M. 1 986. Cell surface glycoconjugates and carbohydrate-bind ing proteins: Possible recognition signals in sensory neurone development. J. Exp. BioI. 1 24: 225-38 Dodd, J., Morton, S. B., Karagogeos, D., Yamamoto, M., Jessell, T. M. 1 988. Spa tial regulation of axonal glycoprotein expression on subsets of embryonic spinal neurons. Neuron I: 1 05-16 Doherty, P., Barton, C. H., Dickson, G., Seaton, P., Rowett, L. H . , e t al. 1 989. Neuronal process outgrowth of human sensory neurons on monolayers of cel ls transfected with cDNAs for five human

involvement in receptor-effector coupling.

Chiu,

A. Y., Matthew, W. D., Patterson,

P. H. 1 986. A monoclonal antibody that blocks the activity of a neurite regener ation-promoting factor. Studies on the binding site and its l oca li zati on in vivo. J. Cell BioI. 103: 1 383-98 Chuang, W., Lagenaur, C. F. 1 990. Central nervous system antigen P84 can serve as a substrate for neurite outgrowth. Dev. BioI. 1 37: 2 1 9-32 Condic, M. L., Bentley, D. 1 989. Removal of the basal lamina in vivo reveals growth cone-basal lamina adhesive interactions and axonal tension in grasshopper embryos. J. Neurosci. 9: 2687-96 Constantine-Paton, M . , Blum, A. S., Men dez-Otero, R., Barnstable, C. J. 1 986. A cell surface molecule distributed in a d or so-ventral gradient in the perinatal rat retina. Nature 324: 459--62 Cooper, J. A. 1 990. The src-family of pro-

factor synthesized in developing skin in

1 52

BIXBY

&

HARRIS

N-CAM isoforms. J. Cell BioI. 109: 78998 Doherty, P., Fruns, M . , Seaton, P., Dickson, G., Barton, C. H . , et al. 1 990. A threshold effect of the major isoforms of NCAM on neurite outgrowth. Nature 343: 464-66 Downward, J., Graves, J. D., Warne, P. H ., Rayter, S., Cantrell, D. A. 1 990. Stimu lation of p21 upon T-cell activation. Nature 346: 7 1 9-23 Edelman, G. M. 1 986. Cell adhesion mol ecules in the regulation of animal form and tissue pattern. Annu. Rev. Cell BioI. 2: 8 1-1 1 6 Edgar, D. 1 989. Neuronal laminin receptors. Trends Neurosci. 1 2: 248-5 1 Edgar, D., Timpl, R., Thoenen, H. 1 988. Structural requirements for the stimu lation of neurite outgrowth by two vari ants of laminin and their inhibition by antibodies. 1. Cell BioI. 106: 1 299- 1 306 Edwards, R. H., Rutter, W. 1., Hanahan, D. 1 989. Directed expression of NGF to pancreatic p cells in transgenic mice leads


,as

to selective hyperinnervation of the islets.

Cell 58: 1 6 1 -70 Ehrig, K., Leivo, I., Argraves, W. S., Ruos lahti, E., Engvall, E. 1 990. Merosin, a tis sue-specific basement membrane protein, is a laminin-like protein. Proc. Natl. A cad. Sci. USA 87: 3264-68 Elkins, T., Hortsch, M., Bieber, A. J., Snow, P. M . , Goodman, C. S. 1 990. Drosophila fascic1in I is a novel homophilic adhesion molecule that along with fasciclin III can mediate cell sorting. 1. Cell BioI. 1 10: UQ5-32 Elkins, T., Zinn, K . , McAllister, L., Bennett, R., Hoffman, M., Goodman, C. S. 1 990. Genetic analysis of a Drosophila neural

cell adhesion molecule: Interaction of

fasciclin I and Abelson tyrosine kinase mutations. Cell 60: 565-75 Faissner, A., Kruse, 1. 1 990. J 1/tenascin is a repulsive substrate for central nervous system neurons. Neuron 5: 627-37 Filbin, M . T., Walsh, F. S., Trapp, B. D., Pizzey, J. A., Tennekoon, G. 1. 1 988. Role of myelin Po protein as a homophilic adhesion molecule. Nature 344: 871-72 Forscher, P., Kaczmarek, L. K., Buchanan, J., Smith, S. J. 1 987. Cyclic AMP induces changes in distribution and transport of organelles within growth cones of Aplysia bag cell neurons. 1. Neurosci. 7: 3600II Fraser, S. E., Carhart, M . S., Murray, B . A., Chuong, C. M., Edelman, C. M. 1 988. Alterations in the Xenopus retinotectal projection by antibodies to Xenopus N CAM. Dev. BioI. 1 29: 2 1 7-30 Fraser. S. E., Hunt. R. K. 1 980. Retinotectal specificity: Models and experiments in

search of a mapping function. Annu. Rev. Neurosci. 3: 3 1 9-52 Furley, A. J., Morton, S. B., M analo, D., Karagogeos, D., Dodd, J., Jessell, T. 1 990. The axonal glycoprotein TAG-I is an im munoglobulin superfamily member with neurite outgrowth-promoting activity. Cell 8 1 : 1 57-70 Geiger, B., Asvnur, Z., Volberg, T., Volk, T. 1 985. Molecular domains of adherens functions. In The Cell in Contact, ed. G . M. Edelman, J. P . Thiery, pp. 46 1 -90. New York: Neuroscience Institute Gennarini, G., Cibelli, G., Rougon, G., Mattei, M.-G., Goridis, C. 1 989. The mouse neuronal cell surface protein F3: A phosphatidylinositol-anchored member of the immunoglobulin superfamily re lated to chicken contactin. J. Cell BioI. 109: 775-88 Gertler, F. B., Bennett, R. L., Clark, M . J . , Hoffmann, F. M . 1 989. Drosophila abl tyrosine kinase in embryonic CNS axons: A role in axonogenesis is revealed through dosage-sensitive interactions with dis abled. Cell 58: 103- 1 3 Gierer, A . 1 987. Directional cues for growing axons in the retinotectal projection. Development 1 0 1 : 479-89 Goldberg, J. I., Kater, S. B. 1 989. Expression and function of the neurotransmitter sero tonin during development of the Helisoma nervous system. Dev. Bioi. 1 3 1 : 483-95 Goodman, C. S., Bastiani, M . J., Doe, C. Q., Dulac, S. 1 986. Growth cone guidance and cell recognition in insect embryos. Dev. Bioi. 3: 283-300 Greene, L. A., Drexler, S. A., Connolly, J. L., Rukenstein, A., Green, S. H. 1 986. Selective inhibition of responses to nerve growth factor and of microtubule-associ ated protein phosphorylation by acti vators of adenylate cyclase. 1. Cell BioI. 103: 1 967-78 Greene, L. A., Shooter, E. M . 1 980. The nerve growth factor. Annu. Rev. Neurosci. 3: 353-405 Greene, L. A., Tischler, A. S. 1 982. PC1 2 pheochromocytoma cultures i n neuro biological research. In Advances in Cellu lar Neurobiology, Vol. 3, ed. S. Federoff, L. Hertz. 3:373-41 4. New York: Aca demic Grumet, M . , Edelman, G. M. 1 988. Neuron glia cell adhesion molecule interacts with neurons and astroglia via different binding mechanisms. J. Cell Bioi. 1 06: 487 503 Gunderson, R. W. 1 987. Response of sensory neurites and growth cones to patterned substrate of laminin and fibro nectin in vitro. Dev. Bioi. 1 2 1 : 423-3 1 Gunderson, R. W., Barrett, J. N. 1 980. Characterization of the turning response -



of dorsal root ganglion neurites toward nerve growth factor. J. Cell Bioi. 87: 54654 Hagag, N., Halegoua, S., Viola, M. 1 986. Inhibition of growth factor-induced differentiation of PCI 2 cells by micro injection of antibody to ras p2 1 . Nature 3 1 9: 680-86 Halfter, W., Deiss, S. 1 984. Axon growth in embryonic chick and quail retina whole mounts in vitro. Dev. Bioi. 1 02: 344-55 Hall, D. E., Neugebauer, K. M ., Reichardt, L. F. 1 987. Embryonic neural retinal cell response to extracellular matrix proteins: Developmental changes and effects of the CSAT antibody. 1. Cell Bioi. 1 04: 623-34 Hama, T., Huang, K. P ., Guroff, G. 1 986. Protein kinase C as a component of a nerve growth factor-sensitive phos phorylation system in PCI 2 cells. Proc. Natl. Acad. Sci. USA 83: 2353-57 Hammarback, J. A., Letourneau, P. C. 1986. Neurite extension across regions of low cell-substratum adhesivity: Implications for the guidepost hypothesis of axonal pathfinding. Dev. Bioi. 1 1 7: 655-62 Hammarback, J. A., McCarthy, J. B., Palm, S. L., Furcht, L. T., Letourneau, P . C. 1 988. Growth cone guidance by substrate bound laminin pathways is correlated with neuron-to-pathway adhesivity. Dev. Bioi. 1 26: 29-39 Hantaz-Ambroise, D., Vigny, M., Koenig, J. 1 987. Heparin sulfate proteoglycan and laminin mediate two different types of neurite outgrowth. J. Neurosci. 7: 22932304 Harrelson, A. L., Goodman, C. S. 1 988. Growth cone guidance in insects: Fascic\in II is a member of the immunoglobulin superfamily. Science 242: 700-8 Harris, W. A. 1 989. Local positional cues in the neuroepithelium guide retinal axons in embryonic Xenopus brain. Nature 339: 2 1 8-21 Harris, W. A., Holt, C. E., Bonhoeffer, F. 1987. Retinal axons with and without their somata, growing to and arborizing in the tectum of Xenopus embryos: A time-lapse video study of single fibres in vivo. Development 1 0 1 : 1 23-33 Hashimoto, S. 1 988. K252a, a potent protein kinase inhibitor, blocks nerve growth factor-induced neurite outgrowth and changes in the phosphorylation of PCI 2 cells. J. Cell Bioi. 1 07: 1 53 1-39 Hatten, M. E., Lynch, M., Rydel, R. E., Sanchez, J., Joseph-Silverstein, J., et al. 1 988. In vitro neurite extension by granule neurons is dependent upon astroglial derived fibroblast growth factor. Dev. Bioi. 1 25: 280-89 Hedgecock, E., Hall, D. 1 990. Homologies

1 53

in the neurogenesis of nematodes, arthro pods and chordates. Semin. Neurosci. 2: 1 59-72 Heffner, C. D., Lumsden, A. G., O'Leary, D. D. 1 990. Target control of collateral extension and directional axon growth in the mammalian brain. Science 247: 2 1 7-20 Heidemann, S. R., Lamoureux, P . , Buxbaum, R. E. 1 990. Growth cone behavior and production of traction force. J. Cell Bioi. I l l : 1 949-57 Hempstead, B. L., Schleifer, L. S., Chao, M . V . 1 989. Expression o f functional nerve growth factor receptors after gene trans fer. Science 243: 373-75 Herndon, M . E., Lander, A. D. 1 990. A diverse set of developmentally regulated proteoglycans is expressed in the rat cen tral nervous system. Neuron 4: 949-61 Holt, C. E .. Garlick, N., Cornel, E. 1990. Targeted lipofection of cDNAs in the embryonic CNS. Neuron 4: 203-14 Huang, K . - P . 1989. The mechanism of pro tein kinase C activation. Trends Neurosci. 1 2 : 425-3 1 Humphries, M . J., Akiyama, S. K . , Komo riya, A., Olden, K., Yamada, K. 1988. Neurite extension of chicken peripheral nervous system neurons on fibronectin: Relative importance of specific adhesion sites in the central cell binding domain and the alternatively spliced type III con necting segment. J. Cell Bioi. 1 06: 1 28997 Hunter, D. D., Shah, V., MerJie, J. P ., Sanes, J. R. 1 989. A laminin-like adhesive protein concentrated in the synaptic cleft of the neuromuscular junction. Nature 338: 22934 Hynes, R. O. 1 987. Integrins: a family of cell surface receptors. Cell 48: 549-54 Igarashi, M., Saito, S., Komiya, Y. 1 990. Vinculin is one of the major endogenous substrates [or intrinsic tyrosine kinases in neuronal growth cones isolated from the fetal rat brain. Eur. J. Biochem. 1 93: 5 5 1 58 Ivins, J., Raper, J., Pittman, R. 1 990. Cal cium levels do not change during contact mediated growth cone collapse. Neurosci. Abs. 1 6: 1 009 Jay, D. G., Keshishian, H. 1990. Laser inac tivation of fasciclin-I disrupts axon adhesion of grasshopper pioneer neurons. Nature 348: 548-50 JesseJl, T. M. 1 988. Adhesion molecules and the hierarchy of neural development. Neuron I : 3-1 3 JesseJl, T., Hynes, M . , Dodd, J . 1990. Carbo hydrates and carbohydrate-binding pro teins in the nervous system. Annu. Rev. Neurosci. 1 3 : 227-55 Johnson, P. W., Abramow-Newerly, W.,


1 54

BIXBY & HARRIS

Seilheimer, B., Sadoul, R., Tropak, M. B., et al. 1989. Recombinant myelin-associ ated glycoprotein confers neural adhesion and neurite outgrowth function. Neuron 3: 377-85 Kadmon, G., Kowitz, A., Altevogt, P., Schachner, M. 1 990. The neural cell adhesion molecule NCAM enhances L l dependent cell-cell interactions. J . Cell BioI. 1 10: 1 93-208 Kaplan, D. R., Martin-Zanca, D., Parada, L. F. 1 99 1 . Tyrosine phosphorylation and tyrosine kinase activity of the trk proto oncogene product induced by NGF. Nature 350: 1 58--60 Karagogeos, D., Morton, S., Casano, F., Dodd, J., Jessell, T. 1 99 1 . Developmental expression of the axonal glycoprotein TAG- I : Differential regulation by central and peripheral neurons in vitro. Develop ment. In press Kapfhammer, 1. P., Raper, 1. A. 1 987a. Col lapse of growth cone structure on contact with specific neurites in culture. J. Neuro sci. 7: 20 1-12 Kapfhammer, 1. P., Raper, J. A. 1 987b. Interactions between growth cones and neurites growing from different neural tis sues in culture. J. Neurosci. 7: 1 595-1600 Kater, S. B., Letourneau, P. c., eds. 1 985. Biology of the Nerve Growth Cone. New York: Liss. 351 pp. Key, B., Akeson, R . A. 1 990. Olfactory neurons express a unique glycosylated form of the neural cell adhesion molecule (N-CAM). J. Cell Bioi. 1 10: 1 729-43 Klein, G., Langegger, M . , Timpl, R., Ekblom, P. 1988. Role of laminin A chain in the development of epithelial cell polarity. Ce/l 55: 33 1-41 Kleitman, N., Simon, D., Schachner, M . , Bunge, R. 1988. Growth of embryonic retinal neurites elicited by contact with Schwann cell surfaces is blocked by anti bodies to L I . Exp. Neurol. 1 02: 298-306 Kligman, D. 1982. Isolation of a protein from bovine brain which promotes neurite extension from chick embryo cerebral cor tex neurons in defined medium. Br. Res. . 250: 93-100 Kligman, D., Marshak, D. 1 985. Purification and characterization of a neurite extension factor from bovine brain. Proc Natl. Acad. Sci. USA 82: 7 1 36--3 9 Koizumi, S., Contreras, M ., Matsuda, Y., Hama, T., Lazarovici, P., Guroff, G. 1988. K252a: a specific inhibitor of the action of nerve growth factor on PCI 2 cells. J. Neurosci. 8: 71 5-21 Kornblihtt, A. R., Umezawa, K., Vibe Pedersen, K., Baralle, F. E. 1 985. Primary structure of human fibronectin: Differ ential splicing may generate at least 1 0 .

polypeptides from a single gene. E MB O 1. 4: 1 755-59 Kypta, R. M . , Goldberg, Y., Ulug, E. T., Courtneidge, S. A. 1990. Association between the PDGF receptor and members of the src family of tyrosine kinases. Cell 62: 48 1 -92 Lagenauer, C., Lemmon, V. 1987. A L I -like molecule, the 8D9 antigen, is a potent sub strate for neurite extension. Proc. Natl. Acad. Sci. USA 84: 7753-57 Lander, A. D. 1987. M olecules that make axons grow. Mol. Neurobiol. I : 2 1 3-45 Lander, A. D. 1989. Understanding the mol ecules of neural cell contacts: Emerging patterns of structure and function. Trends Neurosci. 12: 1 89-95 Landmesser, L. 1 988. Peripheral guidance cues and the formation of specific motor projections in the chick. In From Mes sage to Mind, ed. S. Easter, K . Barald, B. Carlson, pp. 1 2 1-33. Sunderland, Mass: Sinauer Landmesser, L., Dahm, L., Schultz, K., Rutishauser, U . 1 988. Distinct roles for adhesion molecules during innervation of embryonic chick muscle. Dev. Bioi. 1 30: 45-70 Landmesser, L., Dahm, L., Tang, 1. c., Rutishauser, U. 1 990. Polysialic acid as a regulator of intramuscular nerve branch ing during embryonic development. Neu ron 4: 655--67 Lankford, K. L ., DeMello, F. G., Klein, W. L. 1 988. D I-type dopamine receptors inhibit growth cone motility in cultured retina neurons: Evidence that neuro transmitters act as morphogenetic growth regulators in the developing central ner vous system. Proc. Natl. Acad. Sci. USA 85: 2839-43 Large, T. H . , Bodary, S. C., Clegg, D. O., Weskamp, G., Otten, U., Reichardt, L. F. 1986. Nerve growth factor gene expression in the developing rat brain. Science 234: 352-55 Lawler, J., Hynes, R. O. 1986. The structure of human thrombospondin, an adhesive glycoprotein with multiple calcium-bind ing sites and homologies with several different proteins. J. Cell BioI. 1 03: 1 6 3 548 Lefcort, F., Bentley, D. 1987. Pathfinding by pioneer neurons in isolated, opened and mesoderm-free limb buds of embryonic grasshoppers. Dev. Bioi. 1 19: 466-80 Leibrock, 1., Lottspeich, F., Hohn, A., Hofer, M., Hengerer, B., et al. 1 989. Molecular cloning and expression of brain-derived neurotrophic factor. Nature 341 : 1 49-52 Leifer, D . , Lipton, S. A., Barnstable, C. J . , Masland, R. H . 1 984. Monoclonal anti-



1 55

body to Thy-I enhances regeneration of processes by rat retinal ganglion cells in culture. Science 224: 303-6 Lemmon, V., Elmslie, G. J., Hlavin, M. L. 1990. Substrate adhesiveness does not cor relate with neurite growth and fascicu lation. A bstr. Soc. Neurosci. 1 6: 3 1 4 Lemmon, V., Farr, K. L., Lagenauer, C. 1989. L l-mediated axon growth occurs via a homophilic binding mechanism. Neuron 2: 1 597-1 603 Letourneau, P. C. 1975. Cell-to-substratum adhesion and guidance of axonal elon gation. Dev. Bioi. 44: 92- 1 0 1 Letourneau, P . C . 1982. Nerve fiber growth and its regulation by extrinsic factors. I n Neuronal Development, ed. N. C . Spitzer, pp. 2 1 3-54. New York: Plenum Letourneau, P. C., Shattuck, T. A., Ressler, A. H. 1 987. "Pull" and "push" in neurite elongation: Observations on the effects of different concentrations of cytochalasin B and taxo!. Cell Molil. Cytoskeleton 8: 193209 Lindsay, R. M. 1988. Nerve growth factors (NGF, BDNF) enhance axonal regen eration but are n ot required for survival of adult sensory neurons. J. Neurosci. 8: 2394-2405 Lockerbie, R. 0. 1987. The neuronal growth cone: a review of its lomotory, navi gational and target recognition capa bilities. Neuroscience 20: 7 1 9-21 M aher, P. A. 1988. Nerve growth factor induces protein-tyrosine phosphorylation. Proc. Nat!. Acad. Sci. USA 85: 6788-91 Maisonpierre, P., Belluscio, L., Squinto, S., Ip, N., Furth, M . 1990. Neurotrophin-3: A novel factor related to NGF and BDNF . Science 247: 1 446-5 1 M anthorpe, M., Engvall, E., Varon, S . , Muir, D. 1990. Merosin, A laminin-like

1988. Dopamine and serotonin inhi bition of neurite elongation of different identified neurons. J. Neurosci. Res. 1 9: 1 9-26 McKenna, M. P., Raper, J. A. 1988. Growth cone behavior on gradients of substratum bound laminin. Dev. Bioi. 130: 232-36 McLoon, S. C. 1 99 1 . A monoclonal anti body that distinguishes between temporal and nasal retinal halves. J. Neurosci. In press Menesini-Chen, M. G., Chen, J. S., Levi M ontalcini, R. 1978. Sympathetic nerve fibers ingrowth in the central nervous sys tem of neonatal rodent upon intracerebral NGF injections. Arch. Ital. Bioi. 1 1 6: 5384 Moos, M . , Tacke, R., Scherer, H . , Teplow, D., Fruh, K., Schachner, M. 1988. Neural adhesion molecule L l as a member of the immunoglobulin superfamily with bind ing domains similar to fibroncctin. Nature 334: 701-3 Murrain, M., Murphy, A. D., Mills, L. R., Kater, S. B. 1990. Neuron-specific modu lation by serotonin of regenerative out growth and intracellular calcium within the CNS of helisoma-trivolvis. J. Neuro biul. 2 1 : 6 1 1 - 1 8 Neugebauer, K. M., Emmett, C . J., Venstrom, K. A., Reichardt, L. F. 1 99 1 . Vitronectin and thrombospondin pro mote retinal neurite outgrowth: Develop mental regulation and role of i ntegri ns . Neuron 6: 345-58 Neugebauer, K. M., Tomaselli, K. J., Lilicn, J., Reichardt, L. F. 1 988. N-cadherin, NCAM, and integrins promote retinal neurite outgrowth on astrocytes in vitro. J. Cell Bioi. 1 07: 1 1 77-89 Nirenberg, M . , Wilson, S. P., Higashida, H . , Rotter, A., Krueger, K., et a!. 1983. Syn

neurite outgrowth. Abstr. Soc. Neurosci. 1 6: 3 1 4 Matsunaga, M . , Hatta, K . , Nagafuchi, A., Takeichi, M. 1 988. Guidance of optic nerve fibers by N-cadherin adhesion mol ecules. Nature 334: 62-64 Matten, W. T., Aubry, M . , West , J., Maness, P. F. 1 990. Tubulin is phosphorylated at tyrosine by pp60c"", in nerve growth cone membranes. J. Cell Bio/. I l l : 1 959-70 Mattson, M. P., Kater, S. B. 1987. Calcium regulation of neurite elongation and growth cone motility. J. Neurosci. 7: 4034-43 Mattson, M . P., Taylor-Hunter, A., Kater, S. B. 1988. Neurite outgrowth in in dividual neurons of a neuronal popula tion is differentially regulated by calcium and cyclic AMP. J. Neurosci. 8: 1 704-1 I McCobb, D. P., Haydon, P. G., Kater, S.

cells. Cold Spring Harbor Symp. Quant. Bioi. 48: 707- 1 5 Nishizuka, Y . 1988. The molecular hetero geneity of protein kinase C and its impli cations for cellular regulation. Nature 334: 66 1-65 Noda, M . , Ko, M . , Ogura, A., Liu, D., Amano, T., Takano, T. 1985. Sarcoma viruses carrying ras oncogenes induce differentiation-associated properties in a neuronal cell line. Nature 3 1 8: 73-75 Nose, A., Tsuji, K., Takeichi, M . 1 990. Localization of specificity determining sites in cadherin cell adhesion molecules. Cell 6 1 : 1 47-55 O'Connor, T. P., Duerr, J. S., Bentley, D. 1 990. Pioneer growth cone steering decisions mediated by single filopodial contacts in situ. J. Neurosci. 1 0: 3935-46 O'Shea, K. S., Dixit, V. M. 1988. Unique

extracellular matrix molecule, stimulates

B.

apse formation by neuroblastoma hybrid


1 56

BIXBY

&

HARRIS

distribution of the extracellular matrix component thrombospondin in the developing mouse embryo. J. Cell Bioi. 1 07: 2737-48 Osterhout, D. J., Higgins, D. 1 990. Throm bospondin promotes axonal growth in sympathetic neurons. Abstr. Soc. Neuro sci. 1 6: 3 1 2 Patel, N . H., Schafer, B., Goodman, C . S., Holmgren, R. 1 989. The role of segment polarity genes during Drosophila neuro genesis. Genes Dev. 3: 890-904 Patel, N. H., Snow, P. M . , Goodman, C. S. 1 987. Characterization and cloning of fascic1in III: A glycoprotein expressed on a subset of neurons and axon pathways in Drosophila. Cell 48: 975-88 Pittman, R., Ivins, J., Buettner, H. 1 989. Neuronal plasminogen activators-Cell surface binding sites and involvement in neurite outgrowth. J. Neurosci. 9: 426986 Placzek, M., Tessier-Lavigne, M., Jessen, T., Dodd, J. 1990. Orientation of com missural axons in vitro in response to a floor plate-derived chemoattractant. De velopment 1 \0: 1 9-30 Rabacchi, S. A., Neve, R. L., Drager, U. C. 1990. A positional marker for the dorsal embryonic retina is homologous to the high-affinity laminin receptor. Develup ment 1 09: 521-3 1 Ranscht, B. 1 988. Sequence of contactin, a 1 30 kD glycoprotein concentrated in areas of interneuronal contact, defines a new member of the immunoglobulin supergene family in the nervous system. J. Cell Bioi. 1 07: 1 56 1 -73 Ranscht, B., Bronner-Fraser, M. 1 99 1 . T cadherin expression alternates with mi grating neural crest cells in the trunk of the avian embryo. Development 1 1 1 : 1 522 Ranscht, B., Dours, M. T. 1 989. Selective expression of a novel cadherin in the path ways of developing motor- and com missural axons. Abstr. Soc. Neurosci. 1 5: 959 Ranscht, B. , Dours-Zimmerman, M . T. 1 99 1 . T-cadherin, a novel member of the cadherin gene family in the nervous system lacks conserved cytoplasmic sequences. Neuron. In press Raper, J. A., Gruenwald, E. B. 1990. Tem poral retinal growth cones collapse on contact with nasal retinal axons. Exp. Neural. 109: 70-74 Raper, J. A., Kapfhammer, J. P. 1 990. The enrichment of a neuronal growth cone col lapsing activity from embryonic chick brain. Neuron 4: 2 1 -29 Rathjen, F. G., Wolff, J. M ., Frank, R., Bon hoeffer, F., Rutishauser, U. 1 987. Mem-

brane glycoproteins involved in neurite fasciculation. J. Cell Bioi. 103: 343-53 Rauvala, H . , Merenmies, J., Pihlaskari, R., Korkolainen, M., Huhtala, M.-L., Panula, P. 1 988. The adhesive and neurite promoting molecule p30: Analysis of the amino-terminal sequence and production of antipeptide antibodies that detect p30 at the surface of neuroblastoma cells and of brain neurons. J. Cell Bioi. 1 07: 22932305 Rauvala, H . , Pihlaskari, R. 1987. Isolation and some characteristics of an adhesive factor of brain that enhances neurite out growth in central neurons. J. Bioi. Chem. 262: 1 6625-35 Recio-Pinto, E., Rechler, M., Ishii, D. N . 1 986. Effects of insulin, insulin-like growth factor-II, and nerve growth factor on neurite formation and survival in cul tured sympathetic and sensory neurons. J. Neurosci. 6: 1 2 1 1-19 Reichardt, L. F., Bixby, J. L., Hall, D. E., Ignatius, M . J., Neugebauer, K . M . , Tomaselli, K. J. 1 989. Integrins a n d cell adhesion molecules: Neuronal receptors that regulate axon growth on extracellular matrices and cell surfaces. Dev. Neurosci. 1 1 : 332-47

Rinderknecht, E., Humbel, R. E. 1 978. Pri mary structure of human insulin-like growth factor-II. Fed. Eur. Biochem. Soc. Lett. 89: 283-86 Rogelj, S., Klagsbrun, M . , Atzmon, R., Kurokawa, M . , Haimovitz, A., et al. 1 989. Basic fibroblast growth factor is an extra cellular matrix component required for supporting the proliferation of vascular endothelial cells and the differentiation of pe 1 2 cells. J. Cell Bioi. 1 09: 823-3 1 Rogers, S. L., Letourneau, P. C., Pech, I. V. 1 989. The role of fibronectin in neural development. Dev. Neurosci. 1 1 : 248-65 Rosenberg, M. B., Friedmann, T., Robert son, R. c., Tuszynski, M., Wolff, 1. A., et al. 1 988. Grafting genetically modified cells to the damaged brain: Restorative effects of NGF expression. Science 242: 1 575-78 Rosenthal, A., Goeddel, D. V., Nguyen, T., Lewis, M . , Shih, A., et al. 1 990. Primary structure and biological activity ofa novel human neurotrophic factor. Neuron 4: 767-73 Ruegg, M. A., Stoekli, E. T., Lauz, R. B., Streit, P., Sonderegger, P. 1 989. A homo log of the axonally secreted protein axon in-I is an integral membrane protein of nerve fiber tracts involved in neurite fasciculation. J. Cell Bioi. \ 09: 2363-68 Ruoslahti, E., Pierschbacher, M . D. 1 987. New perspectives in cell adhesion: RGD and integrins. Science 238: 49 1-97



Rutishauser, U., Acheson, A., Hall, A., Mann, D., Sunshine, 1. 1 988. The neural cell adhesion molecule (NCAM) as a regu lator of cell-cell interactions. Science 240: 53-57 Salzer, 1. L., Holmes, W. P., Colman, D. R. 1 987. The amino acid sequences of the myelin-associated glycoproteins: hom ology to the immunoglobulin gene super family. J. Cell Bioi. 1 04: 957--65 Schlosshauer, B., Schwarz, U., Rutishauser, U. 1 984. Topological distribution of different forms of neural cell adhesion molecule in the developing visual system . Nature 3 1 0: 1 41 -43 Schubert, D., Lacorbiere, M., Whitlock, c., Stallcup, W. 1 978. Alterations in the sur face properties of cells responsive to nerve growth factor. Nature 273: 7 1 8-23 Schuch, U., Lohse, M. 1., Schachner, M. 1989. Neural cell adhesion molecules influence second messenger systems. Neu ron 3: 1 3-20 Schwab, M . E. 1 990. Myelin-associated inhibitors of neurite growth and regen eration in the CNS. Trendv Neurosci. 1 3: 452-56 Seilheimer, B., Schachner, M. 1988. Studies of adhesion molecules mediating inter actions between cells of peripheral nervous system indicate a major role for L1 in mediating sensory neuron growth on Schwann cells in culture. J. Cell Bioi. 1 07: 341 -5 1 Shaw, L . M ., Messier, J . M . , Mercurio, A. M. 1 990. The activation-dependent adhesion of macrophages to laminin involves cytoskeletal anchoring and phos phorylation of the (l.6fJ 1 intcgrin. J. Cell Bioi. 1 10: 2 167-74 Silver, J., Poston, M . , Rutishauser, U. 1 987. Axon pathway boundaries in the develop ing brain. I. Cellular and molecular deter minants that separate the optic and olfac tory projections. J. Neurosci. 7: 226472 Silver, 1., Rutishauser, U. 1 984. Guidance of optic axons in vivo by a performed adhesive pathway on neuroepithelial end feet. Dev. Bioi. 1 06: 485-99 Silver, R. A., Lamb, A. G., Bolsover, S. R. 1 989. Elevated cytosolic calcium in the growth cone inhibits neurite elongation in

neuroblastoma cells: Correlation of behavioral states with cytoplasmic cal cium concentration. J. Neurosci. 9: 400720 Silver, R. A., Lamb, A. G., Bolsover, S. R . 1 990. Calcium hotspots caused by L-chan nel clustering promote morphological changes in neuronal growth cones. Nature 343: 751-54 Skene, 1. H. P . 1 989. Axonal growth associ-

1 57

ated proteins. Annu. Rev. Neurosci. 1 2: 1 27-56 Smith, S. J. 1988. Neuronal cytomechanics: the actin-based motility of growth cones. Science 242: 708-1 5 Snow, P . M . , Bieber, A. 1., Goodman, C . S. 1 989. Fasciclin III: A novel homophilic adhesion molecule in Drosophila. Cell 59: 3 1 3-23 Snow, P. M., Lemmon, V., Carrino, D. A., Caplan, A. I., Silver, 1. 1 990. Sulfated pro teoglycans in astroglial barriers inhibit neurite outgrowth in vitro. Exp. Neurol. 1 09: 1 1 1-30 Sonnenberg, A., Linders, C. J. T., Modder man, P. W., Damsky, C. H., Aumailley, M., Timpl, R. 1 990. Integrin recognition of different cell-binding fragments oflami nin (PI , E3, E8) and evidence that (l.6{3 1 but not (l.6fJ4 functions as a major receptor for fragment E8. J. Cell BioI. 1 10: 2 1 4555 Speidel, C. C. 1 933. Studies of living nerves. II. Activities of amoeboid growth cones, sheath cells, and myelin segments, as revealed by prolonged observation of indi vidual nerve fibers in frog tadpoles. Am. J. Anal. 52: 1-75 Sperry, R. W. 1 963. Chemoaffinity in the orderly growth of nerve fiber patterns and connections. Proc. Natl. Acad. Sci. USA 50: 703-1 0 Spring, J., Beck, K . , Chiquet-Ehrismann, R . 1 989. Two contrary functions of tenascin: dissection of the active sites by recom binant tenascin fragments. Cell 59: 32534 Springer, 1. E. 1 988. Nerve growth factor receptors in the central nervous system. Exp. Neurol. 1 02: 354--65 Springer, T. A. 1 990. Adhesion receptors of the immune system. Nature 346: 42534 Stahl, B., Muller, B., von Boxberg, Y., Cox, E. c., Bonhoelfer, F. 1 990. Biochemical characterization of a putative axonal guid ance molecule of the chick visual system. Neuron 5: 735-43 Steindler, D. A" O'Brien, T. F., Laywell, E., Harrington, K., Faissner, A., Schachner, M. 1 990. Boundaries during normal and abnormal brain development: In vivo and in vitro studies of glia and glycoconjugates. Exp. Neurol. 1 09: 35-56 Stern, C. D., Norris, W. E., Bronner-Fraser, M., Carlson, G. 1., Faissner, A., et al. 1989. l 1ftenascin-related molecules are not responsible for the segmented pattern of neural crest cells or motor axons in the chick embryo. Development 1 07: 309-1 9 Sternweis, P . C . , Robishaw, 1 . D . 1 984. Iso lation of two proteins with high affinity for guanine nucleotides from membranes


1 58

BIXBY

& HARRIS

of bovine brain. 1. Bioi. Chern. 259: 1380613 Stoekli, E. T., Kuhn, T. B., Duv, C O., Ruegg, M. A., Sonderegger, P. 1 99 1 . The axon ally secreted protein axonin-I is a potent substratum for neurite growth. 1. Cell Bioi. 1 12: 449-55 Strittmatter, S. M ., Valenzuela, D., Kennedy, T. E., Neer, E. 1., Fishman, M. C. 1 990. Go is a major growth cone protein subject to regulation by GAP-43. NalUre 344: 836-4 1 Suarez-Isla, B. A., Pelto, D. 1., Thompson, J. M., Rapoport, S. I. 1 984. Blockers of calcium permeability inhibit neurite exten sion and formation of neuromuscular syn apses in cell culture. Dev. Brain Res. 1 4: 263-70 Suzue, T., Kaprielian, Z., Patterson, P. H . 1990. A monoclonal antibody that defines rostrocaudal gradients in the mammalian nervous system. Neuron 5: 42 1-3 1 Takeichi, M. 1990. Cadherins: a molecular family important in selective cell adhesion. Annu. Rev. Biochern. 59: 237-52 Tapley, P., Horwitz, A., Buck, C., Duggan, K., Rohrschneider, L. 1 989. Integrins iso lated from Rous sarcoma virus-trans formed chicken embryo fibroblasts. Onco gene 4: 325-33 Taraboletti, G., Roberts, D. D., Liotta, L. 1 987. Thrombospondin-induced tumor cell migration: Haptotaxis and chemotaxis are mediated by different molecular domains. J. Cell Bioi. 1 05: 2409-1 5 Tessier-Lavigne, M . , Placzek, M . , Lumsden, A. G., Dodd, J., Jessell, T. M. 1988. Chemotropic guidance of developing axons in the mammalian central nervous system. Nature 336: 775-78 Tomaselli, K. 1., Hall, D. E., Flier, L. A., Gehlsen, K. R., Turner, D. C, et al. 1 990. A neuronal cell line (PCI2) expresses two P I-class integrins-IX I P I and IXJP I-that recognize different neurite outgrowth promoting domains in laminin. Neuron 5: 65 1-62 Tomaselli, K. 1., Neugebauer, K. N., Bixby, J. L., Lilien, J., Reichardt, L. F. 1 988. N cadherin and integrins: two receptor sys tems that mediate neurite outgrowth on astrocyte surfaces. Neuron 1: 33-43 Tomaselli, K. J., Reichardt, L. F., Bixby, J. L. 1 986. Distinct molecular interactions mediate neuronal process outgrowth on non-neuronal cell surfaces and extra cellular matrices. 1. Cell Bioi. 1 03: 265972 Traynor, A. E., Schubert, D. 1 984. Phos pholipases elevate cyclic AMP levels and promote neurite extension in a clonal nerve cell line. Dev. Br. Res. 14: 1 97-204 Trisler, D. 1 990. Cell recognition and pattern

formation in the developing nervous system. 1. Exp. Bioi. 1 53: 1 1-27 Trisler, D., Bekenstein, J., Daniels, M . P. 1986. Antibody to a molecular marker of cell position inhibits synapse formation in the retina. Proc. Natl. Acad. Sci. USA 83: 41 94-98 Trisler, D., Collins, F. 1 987. Corresponding spatial gradients of TOP molecules in the developing retina and optic tectum. Sci ence 237: 1 208-9 Trisler, G., Schneider, M . D., Nirenberg, M . 1 98 1 . A topographic gradient o fmolecules in retina can be used to identify neuron position. Proc. Natl. A cad. Sci. USA 78: 2 145-49 Unsicker, K., Reichert-Preibsh, H., Schmidt, R., Pettmann, B., Labourdette, G., Sen sen brenner, M. 1 987. Astroglial and fibroblast growth factors have neuro trophic functions for cultured peripheral and central nervous system neurons. Proc. Natl. Acad. Sci. USA 84: 5459-63 Van Eldik, L. J., Staecker, J. L., Win ningham-Major, F. 1 988. Synthesis and expression of a gene coding for the calcium-modulated protein S l OOP and designed for cassette-based, site-directed mutagenesis. 1. Bioi. Chern. 263: 783037 Vielmetter, 1., Stolze, B . , Bonhoeffer, F., Stuermer, C. A. 1 990. In vitro assay to test differential substrate affinities of growing axons and migratory cells. Exp. Brain Res. 8 1 : 283-87 Walter, 1., Allsopp, T. E., Bonhoeffer, F . 1 990. A common denominator of growth cone guidance and collapse. Trends Neuro sci. 1 3: 447-52 Wayner, E. A., Carter, W. G. 1 987. Identi fication of multiple cell surface receptors for fibronectin and type VI collagen. 1. Cell Bioi. 1 05 : 1 873-84 Werb, Z., Tremble, P. M . , Behrendtsen, 0., Crowley, E., Damsky, C. H . 1 989. Signal transduction through the fibronectin receptor induces collagenase and stro melysin gene expression. J. Cell BioI. 109: 877-89 Wessells, N. K., Nuttall, R. P. 1 978. Normal branching, induced branching, and steer ing of cultured parasympathetic motor neurons. Expl. Cell Res. 1 15: 1 1 1-22 Wewer, U. M . , Taraboletti, G" Sobel, M . E., Albrechtsen, R . , Liotta, L. A. 1 987. Role of laminin receptor in tumor cell migration. Cancer Res. 47: 5691 -98 Williams, A. F., Barclay, A. N. 1988. The immunoglobulin superfamily: Domains for cell surface recognition. Annu. Rev. Imrnunol. 6: 381-405 Winningham-Major, F., Staecker, 1. L., Barger, S. W., Coats, S., Van Eldik, L.


J. 1 989. Neurite extension and neuronal survival activities of recombinant S l OOp proteins that differ in the content and posi tion of cysteine residues. J. Cell Bioi. 1 09:


3063-7 1

Wu, B.-Y., Fodor, E. J. B., Edwards, R. H . , Rutter, W. J . 1 989. Nerve growth factor induces the proto-oncogene c-jun in PC1 2 cells. J. Bioi. Chern. 264: 9000-3 Wyatt, S., Shooter, E. M . , Davies, A. M . 1 990. Expression o f the NGF receptor gene in sensory neurons and their cutaneous targets prior to and during innervation. Neuron 4: 42 1-27 Zinn, K., McAllister, L., Goodman, C. S. 1 988. Sequence analysis and neuronal

1 59

expression of fasciclin I in grasshopper and Drosophila. Cell 53: 577-87 Zipser, B., Morell, R., Bajt, M . L. 1 989. Defasciculation as a neuronal pathfinding strategy: Involvement of a specific glyco protein. Neuron 3: 62 1 -30

NOTE ADDED IN PROOF

Burgoon, M . P., Grumet, M., Mauro, V., Edelman, G. M . , Cunningham, B . A. 1 99 1 . Structure of the chicken neuron-glia adhesion molecule Ng-CAM: Origin of the polypeptides and relation to the Ig superfamily. J. Cell Bioi. 1 1 2: 1 0 1 7-29

Molecular mechanisms of scar-sourced axon growth inhibitors.

Signaling mechanisms of axon guidance and early synaptogenesis.

Axon guidance: FLRTing promotes attraction.

Celsr3 and Fzd3 in axon guidance.

Microtechnologies for studying the role of mechanics in axon growth and guidance.

Where does axon guidance lead us?

Drosophila Hrp48 Is Required for Mushroom Body Axon Growth, Branching and Guidance.

Independent signaling by Drosophila insulin receptor for axon guidance and growth.

Hsc70 chaperone activity underlies Trio GEF function in axon growth and guidance induced by netrin-1.

Axon guidance proteins in neurological disorders.

Mitochondrial Dynamics in Retinal Ganglion Cell Axon Regeneration and Growth Cone Guidance.

A mathematical model explains saturating axon guidance responses to molecular gradients.

Coevolution of axon guidance molecule Slit and its receptor Robo.

Using Xenopus laevis retinal and spinal neurons to study mechanisms of axon guidance in vivo and in vitro.

Partial interchangeability of Fz3 and Fz6 in tissue polarity signaling for epithelial orientation and axon growth and guidance.

Glypican Is a Modulator of Netrin-Mediated Axon Guidance.

Morphogenesis of the C. elegans Intestine Involves Axon Guidance Genes.

Netrin signaling and Rac GTPases.

EphA4 receptor shedding regulates spinal motor axon guidance.

Canonical wnt signaling is required for commissural axon guidance.

Axon guidance: Setting up for seeing in slow motion.

Charcot-Marie-Tooth 2b associated Rab7 mutations cause axon growth and guidance defects during vertebrate sensory neuron development.

Filamin and multiple genes involved in axon guidance in Drosophila.

Behavioral mechanisms of mammalian sperm guidance.