differential expression of cell adhesion molecules on ... · differential expression of cell...

DEVELOPMENTAL BIOLOGY 178, 101–112 (1996)ARTICLE NO. 0201

Differential Expression of Cell Adhesion Moleculeson Trigeminal Cutaneous and Muscle Afferents

Sheryl A. Scott, Chi Ai Chau, Kathryn F. Randazzo,Regan Stone, and Sharon M. CahoonDepartment of Neurobiology and Anatomy, University of Utah School of Medicine,50 N. Medical Drive, Salt Lake City, Utah 84132

Previous work from our laboratory (Woodbury and Scott, 1995) suggested that embryonic cutaneous and muscle afferentsmight express different surface cell adhesion molecules (CAMs) on their growth cones. To examine this possibility directlywe measured the relative levels of expression of various adhesion molecules on growth cones of neurons from the dorsomed-ial portion of the trigeminal ganglion (DM-TG), which are largely cutaneous, and from the trigeminal mesencephalicnucleus (TMN), which are exclusively muscle afferents. Axonin-1, L1, BEN, and N-cadherin were expressed more abundantlyon DM-TG growth cones, whereas N-CAM was more abundant on TMN neurons. Expression of polysialated N-CAM wassimilar on the two populations; addition of NGF and NT-3 appeared to increase expression of polysialated N-CAM onTMN neurons. Although the levels of L1 and axonin-1, both of which bind L1, were markedly different on TMN and DM-TG neurons, these differences were not sufficient to cause dramatic differences in the growth rates of TMN and DM-TGneurons on L1. q 1996 Academic Press, Inc.

INTRODUCTION vitro (Woodbury and Scott, 1995). These studies took advan-tage of the anatomical separation of cutaneous and muscleafferents in the avian trigeminal system; sensory neurons

The major function of primary sensory neurons, such as from the dorsomedial portion of the trigeminal ganglionthose in dorsal root ganglia (DRG) or trigeminal ganglia, is (DM-TG) (D’Amico-Martel and Noden, 1983) and the tri-to convey information to the central nervous system about geminal mesencephalic nucleus (TMN) (Narayanan andthe location, nature, and intensity of peripheral stimuli. Narayanan, 1978) are both derived from the neural crest, butDuring embryonic development sensory neurons must, serve primarily cutaneous and exclusively muscle targets,therefore, connect precisely with the appropriate central respectively. Thus, there do indeed appear to be differencesand peripheral targets. A number of studies have shown that between the growth cones of different types of sensory neu-sensory neurons in DRG make both central and peripheral rons, at least at the ages we studied. The present reportconnections with a high degree of precision (reviewed in addresses this problem directly.Scott, 1992; see also Mirnics and Koerber, 1995a,b; Silos- Growth cones of sensory neurons express a wide varietySantiago et al., 1995). An intriguing question, for which we of surface cell adhesion molecules (CAMs) that regulatehave no definitive answer, is whether sensory neurons are their interactions with other axons and the environment,specified or committed to make these connections prior to which can be categorized into three major classes—Ca2/-axon outgrowth and target selection. If so, the growth cones dependent cadherins, Ca2/-independent immunoglobulin-of nascent sensory neurons must make ‘‘informed choices’’ like molecules, and integrin receptors for extracellular ma-as they navigate the terrain to their respective targets, and trix molecules (reviewed in Landmesser, 1994). We selectedthe growth cones of neurons that project to one target, such for study six such molecules that had previously been re-as a particular region of skin, must differ somehow from ported to be either heterogeneously distributed or develop-those that project to a different target, such as a particular mentally regulated on sensory neurons and which had beenmuscle. shown to be involved in promoting adhesion, outgrowth,

We have recently shown that avian cutaneous and muscle or fasciculation of sensory neurons (see Landmesser, 1994,and Discussion for references). Using a technique describedafferents respond differently to several potential targets in

101

0012-1606/96 $18.00Copyright q 1996 by Academic Press, Inc.All rights of reproduction in any form reserved.

AID DB 8304 / 6x11$$$221 07-30-96 09:24:49 dbas AP: Dev Bio

102 Scott et al.

scope slides in glycerol containing p-phenylenediamine (Sigma, P-in abstract form for the analysis of L1 and N-CAM on DRG6001) to prevent photobleaching. DRG were included in initial ex-neurons (Honig, 1991), we found differences in the relativeperiments as a positive control; DRG incubated in PBS containinglevels of expression between trigeminal cutaneous and mus-1% normal goat serum (NGS) instead of primary antibody werecle afferents for five of the six molecules examined. Whenincluded as a negative control in all experiments.our experiments were nearing completion a full report de-

Explants were stained with the following antibodies diluted inscribing the expression of CAMs on DRG neurons appeared PBS with 1% NGS: anti-axonin-1 [Developmental Studies Hybrid-(Honig and Kueter, 1995). The patterns of expression of ad- oma Bank (DSHB), 23.4-5; supernatant diluted 1:1], BEN (DSHB;hesion molecules on cutaneous and muscle afferents in undiluted supernatant), anti-L1 (DSHB, 8D9; supernatant dilutedDRG differ from those reported here for trigeminal cuta- 1:1), anti-N-cadherin (Sigma, C-2542; diluted 1:100), anti-N-CAM

(DSHB, 5e; supernatant diluted 1:1), or anti-polysialated N-CAMneous and muscle afferents. These differences cannot be(DSHB, 5A5, supernatant diluted 1:1). Dr. Urs Rutishauser gener-fully explained by the age of embryos studied or conditionsously supplied additional anti-N-cadherin (R054; diluted to 30 mg/under which neurons were cultured. Together these studiesml), 5e (50 mg/ml), and 5A5 (diluted 1:200). Several of these antigenssuggest that, despite their common neural crest origin, tri-and antibodies are also known by other names. Axonin-1, a membergeminal cutaneous and muscle afferents may not be compa-of the immunoglobulin superfamily, is the avian homologue ofrable to their counterparts in DRG. Some of this work hasmammalian TAG-1 (Zuellig et al., 1992). Axonin-1 has recently

been reported previously in abstract form (Scott et al., 1994). been shown to be identical to SC2 (Sakurai et al., 1994). For simplic-ity, this CAM will be referred to here as axonin-1. BEN (Pourquieet al., 1990) is also a member of the immunoglobulin superfamily.This CAM has been identified and named by several laboratories:METHODSSC-1 (Tanaka and Obata, 1984), DM-GRASP (Burns et al., 1991),and the JC7 antigen (El-Deeb et al., 1992). We will refer to it hereCulturessimply as BEN, since that is the antibody we used. The 8D9 antigen

Explant cultures were established from White Leghorn chicken is an L1-like molecule (Lemmon and McLoon, 1986) in the immu-embryos from fertile eggs provided by a local supplier. The dor- noglobulin superfamily, identical to G4 (Rathjen et al., 1987) andsomedial portion of trigeminal ganglia (DM-TG), the trigeminal Ng-CAM (reviewed in Grumet, 1992). For simplicity, the 8D9 anti-mesencephalic nucleus (TMN), and several lumbosacral dorsal root gen will be referred to here as L1. Antibody 5e recognizes the extra-ganglia (DRG) from Embryonic Day 10 embryos [E10, St. 36 (Ham- cellular portion of all forms of N-CAM, whereas 5A5 recognizesburger and Hamilton, 1951)] were dissected in L15 (GIBCO, Grand only the polysialated form (Dodd et al., 1988).Island, NY) and cut into 6–8 pieces. Generally, explants from two Dissociated cell cultures were fixed, permeabilized with 0.05%ganglia or two TMN were placed on a 22-mm acid-cleaned cov- Triton X-100, and stained overnight at 47C with either 3A10 (DSHB;erglass that had been coated with poly-DL-ornithine (500 ng/ml 0.15 supernatant diluted 1:1–1:5) and/or anti-NAPA-73 (DSHB, E/C8;M borate buffer, pH 8.7; Sigma, P-8638, St. Louis, MO) and laminin undiluted supernatant) to label neurites.[GIBCO; 20 mg/ml phosphate-buffered saline (PBS)] and placed in a35-mm petri dish. Explants were grown for 18–24 hr in definedmedium consisting of F14 supplemented with N2 additives (F14/ ImagingN2; GIBCO), 23 mM NaHCO3, 50 units/ml penicillin, and 50 mg/

The goal of the first series of experiments was to compare theml streptomycin. Cultures of DM-TG and DRG were further sup-expression of different surface antigens on growth cones of DM-TGplemented with 10 ng/ml nerve growth factor (NGF) (Sigma, N-and TMN neurons by measuring the relative intensity of antibody6009). TMN cultures were supplemented with either 10 ng/mllabeling. Images of individual growth cones were viewed with abrain-derived neurotrophic factor (BDNF) or with 10 ng/ml BDNF,1001 objective, captured with an intensified CCD camera (eitherneurotrophin-3 (NT-3), and NGF. BDNF and NT-3 were generouslyPulnix or Attofluor), and recorded on a Panasonic optical memoryprovided by Regeneron Pharmaceuticals, Inc. (Tarrytown, NY). Ex-disk recorder (OMDR). Each growth cone assayed was judged to beplant cultures were established in duplicate.a single growth cone arising from a single neurite; growth conesAdditional cultures were established from TMN and DM-TGarising from apparent fascicles were excluded. Camera settingsfrom St. 29–31 (E61

2–712). In each of these experiments half the cul-

were adjusted to optimize contrast at the beginning of each experi-tures were grown in F14/N2, as described above, and the other halfment and were then kept constant during the rest of the recordingin F14/N2 containing 1% horse serum and 2.5% embryo extractsession; both cameras are linear over the entire measuring range.(cf. Honig and Kueter, 1995). DM-TG neurons were supplementedTo control for nonuniformity in staining across or between slides,with NGF, and TMN neurons with all three neurotrophic factors.as well as to correct for possible drift in the xenon burner or CCDcamera, each slide was divided into four quadrants. Images of ap-proximately 25 randomly selected growth cones were collectedImmunohistochemistryfrom explants in one quadrant on the first slide for each type ofneuron; images of 25 growth cones were then collected from oneMost cultures were fixed for 15–20 min with 4% paraformalde-

hyde in PBS containing 0.48 mM CaCl2, prior to staining. In several quadrant on the second slide for each type of neuron. This sequencewas then repeated with a different quadrant on both slides for eachexperiments live cultures were stained with anti-axonin-1 prior to

fixation. Explants were incubated in a primary antibody for 1 hr at type of neuron. In total, approximately 100 growth cones wereimaged for each type of neuron in most experiments. The intensityroom temperature, followed by 30 min incubation in the appro-

priate FITC-coupled secondary antibody diluted 1:200–1:300 (Cap- of recorded images was then measured using Image-1 software (Uni-versal Imaging Corp., West Chester, PA). To correct for differencespel, Durham, NC). The coverglasses were then mounted on micro-

Copyright q 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.


103Adhesion Molecules on Trigeminal Sensory Neurons

in overall background brightness, a 1-cm square adjacent to each cone in that experiment, which was arbitrarily set as 100%. Toallow comparison among all experiments, only TMN neurons sup-growth cone was imaged and subtracted from the intensity value

obtained for that growth cone. Labeling intensity was then ex- plemented with BDNF or DM-TG neurons were considered in se-lecting the brightest growth cones, although occasionally growthpressed as a percentage of the intensity of the brightest growth

FIG. 1. Representative examples of CAM expression on growth cones of dorsomedial trigeminal (DM-TG) (A, C, E, G, I, K) and trigeminalmesencephalic nucleus (TMN) (B, D, F, H, J, L) neurons from E10 chick embryos. Expression of axonin-1 (A, B), BEN (C, D), L1/8D9 (E,F), and N-cadherin (G, H) was higher on DM-TG than TMN neurons, whereas expression of N-CAM (I, J) was higher on TMN neurons.Expression of PSA-N-CAM (K, L) was similar on the two types of neurons. Bar, 10 mm.



104 Scott et al.

cones from TMN neurons grown in all trophic factors were actually in F14 supplemented with 10% horse serum and appropriate neuro-trophic factors. After approximately 18 hr cultures were fixed,the brightest; the latter, therefore, had intensity values greater than

100%. Differences in labeling were tested for significance using stained with neurofilament antibodies (3A10 or anti-NAPA-73),and measured from recorded images, as described above.the Kruskal–Wallis or Mann–Whitney test and were considered to

be statistically significant for P values equal to or less than 0.02.The images shown in Fig. 1 are conventional 35-mm photographs RESULTS

taken at the same exposure and printed using digital imaging tech-nology. Differential Expression of Adhesion Molecules on

DM-TG and TMN NeuronsThe goal of initial studies was to examine the expressionNeurite Growth

of surface CAMs on the growth cones of sensory neuronsThe growth of DM-TG and TMN neurons on L1 and laminin from the DM-TG, which are primarily cutaneous afferents,

was compared in two ways. and from the TMN, which are entirely muscle afferents,Growth rates. Glass-bottomed wells were coated with 40 ml of

under conditions similar to those of our previous timelapsenitrocellulose (2.5 cm2 dissolved in 5 ml amyl acetate; Schleicher &studies (Woodbury and Scott, 1995). To this end, DM-TGSchuell, Keene, NH) for 15 min; excess nitrocellulose was removedexplants were supplemented with NGF and TMN explantsby aspiration, and the dishes were air-dried. A small spot (5 ml) ofsupplemented with BDNF, and the expression of adhesionmouse L1 (a generous gift of Dr. Carl Lagenaur) was then appliedmolecules was measured (Methods). Under these conditionson one half of the dish. Excess L1 was removed after 10 min and

additional L1 was applied to the same spot. At that time approxi- we observed differences in the relative levels of expressionmately 25–35 ml laminin (50 mg/ml PBS) was applied on the other of all of the adhesion molecules examined, with the excep-half of the dish, as well as to additional control dishes. Excess L1 tion of the polysialated form of N-CAM. Representativeand laminin were removed after 10 min, the spots washed with examples of the labeling observed under these conditionsPBS, and the wells were blocked for at least 30 min with F14 con- are shown in Fig. 1; the results of all experiments are listedtaining 10% horse serum. Ten to twelve small pieces from a single

in Table 1 and summarized in Fig. 2.E10 DM-TG or TMN were then placed in each dish and grownTrigeminal cutaneous and muscle afferents differed mostovernight in F14/N2 supplemented with NGF or BDNF and NT-

markedly in expression of axonin-1. The pattern of labeling3, respectively.in vitro mirrored that previously reported for avian trigemi-The next day the growth of individual neurites was followednal and TMN neurons in vivo (Halfter et al., 1994). Allwith time lapse videomicroscopy. Culture dishes were filled with

warm, equilibrated F14/N2 and additional neurotrophic factors, cutaneous neurites were brightly labeled with antibodies tosealed with a glass slide, placed on the stage of an Olympus IMT2 axonin-1, whereas muscle afferents were unlabeled and didmicroscope, and viewed with a 401 phase objective. The stage not differ significantly from negative controls. This patternwas enclosed in a thermostatically controlled chamber that was was observed consistently; in addition to the three experi-maintained at 377C throughout the recording session. Isolated ments listed in Table 1, two other sets of cultures weregrowth cones that grew without obvious interactions with neigh-

stained with anti-axonin-1, but were not measured sinceboring growth cones were selected for study. Images were capturedthey clearly showed the same pattern.every 30 sec under low light conditions using a Pulnix CCD camera

Cutaneous and muscle afferent growth cones both ex-and recorded on an OMDR. Most growth cones were followed forpressed all of the other adhesion molecules tested, althoughat least 1 hr (average recording time Å 116 { 47 min; range Å 30the average level of expression differed between the twoto 243 min). Neurite length was measured from an arbitrary place

on the image at 10-min intervals using Image-1 software; growth types of neurons. L1, BEN, and N-cadherin were expressedrates were calculated as the total distance grown during the time significantly more abundantly on DM-TG growth cones inobserved. all experiments. In contrast, expression of N-CAM, as ob-

Neuron morphology. In one experiment (KF1, Table 2) cov- served with both antibodies that recognize all forms of N-erglass wells were coated with nitrocellulose and L1 or laminin CAM (5e) as well as antibodies specific for the polysialatedas described above. DM-TG and TMN were dissected from E10

form (5A5), varied among different experiments (Table 1).embryos, dissociated with 0.1% trypsin, and the nonneuronal cellsWe have no explanation for this variability, but similar vari-removed by differential sedimentation in F14 containing 20% horseability in labeling with these antibodies has been reportedserum at 27C (Davies, 1989a). Neurons were plated at low densityby others (Honig and Kueter, 1995). Nevertheless, on aver-in F14/N2 supplemented with NGF (for DM-TG) or BDNF and NT-age N-CAM appeared to be significantly more abundant on3 (for TMN). Approximately 18 hr later cultures were fixed with

4% paraformaldehyde and phase-contrast images of isolated neu- TMN muscle afferents than on cutaneous afferents, as as-rons were recorded on an OMDR. The length of the longest neurite sayed by antibody 5e. In contrast, there was no apparentof each cell was measured using Image-1 software, and the number difference in the amount of PSA-N-CAM expressed by theof branch points was counted. two populations of neurons.

In two experiments (SMC7 and 8; Table 2) 35-mm tissue culturedishes were coated with nitrocellulose dissolved in methanol (La- Effects of Neurotrophic Factors on CAM Expressiongenaur and Lemmon, 1987), treated with single applications of L1

Expression of CAMs on trigeminal sensory neurons dif-or laminin, and washed and blocked, as above. Dissociated E10and/or E11 DM-TG and TMN neurons were plated at low density fered markedly from that reported for sensory neurons in




TABLE 1Relative Intensity of Antibody Labeling of Growth Cones

Antigen (antibody) DM-TG TMN (BDNF)a P TMN (ALL)b Pc

Axonin-1 (23.4–5)CAC27 40.52 { 19.83 (127) 6.92 { 5.11 (93) õ0.0001CAC28 28.34 { 17.62 (126) 3.86 { 2.91 (123) õ0.0001CAC81 40.32 { 19.33 (107) 4.98 { 4.69 (100) õ0.0001 3.55 { 2.52 (100) 0.0400 (NS)

Average 36.20 { 19.74 (360) 5.11 { 4.40 (316) õ0.0001 3.55 { 2.52 (100) 0.0400 (NS)BEN

CAC24 26.94 { 18.14 (100) 12.08 { 6.91 (77) õ0.0001CAC45 30.46 { 14.81 (98) 16.82 { 10.25 (103) õ0.0001CAC46 25.67 { 18.18 (99) 7.77 { 2.83 (102) õ0.0001CAC98 39.92 { 18.37 (107) 14.55 { 5.87 (105) õ0.0001 19.42 { 6.25 (103) õ0.0001CAC100 35.77 { 15.65 (103) 18.83 { 6.69 (107) õ0.0001 18.92 { 6.45 (101) 0.8745 (NS)

Average 31.91 { 19.89 (507) 14.16 { 7.94 (494) õ0.0001 19.17 { 6.34 (204) õ0.0001L1(8D9)

CAC11 57.11 { 21.56 (47) 41.11 { 17.02 (39) 0.0003CAC30 43.93 { 17.94 (136) 25.90 { 9.04 (127) õ0.0001CAC55 51.48 { 19.34 (109) 37.68 { 15.08 (109) õ0.0001CAC73a 69.89 { 14.37 (105) 59.27 { 15.93 (103) õ0.0001 62.57 { 13.24 (100) 0.1944 (NS)

Average 54.43 { 20.56 (397) 39.96 { 18.99 (378) õ0.0001 62.57 { 13.24 (100) 0.1944 (NS)N-cadherin (C-2542)

CAC44 41.48 { 13.16 (100) 14.55 { 4.89 (99) õ0.0001CAC58 41.25 { 16.65 (106) 34.45 { 12.04 (103) 0.0009CAC59 46.01 { 15.01 (129) 38.57 { 11.81 (124) õ0.0001

N-cadherin (R054)CAC86d 42.41 { 18.18 (116) 25.71 { 10.56 (107) õ0.0001 27.04 { 12.93 (123) 0.6109 (NS)

Average 42.96 { 15.97 (451) 28.92 { 13.81 (433) õ0.0001 27.04 { 12.93 (123) 0.6109 (NS)N-CAM (5e)

CAC33 28.27 { 14.95 (116) 42.40 { 15.13 (129) õ0.0001CAC34 39.70 { 17.32 (138) 39.51 { 10.91 (139) 0.3717 (NS)CAC38 46.23 { 17.17 (109) 49.88 { 16.82 (89) 0.1283 (NS)CAC60 33.29 { 10.16 (109) 51.42 { 14.33 (92) õ0.0001CAC83ad 44.92 { 14.92 (89) 60.44 { 16.46 (79) õ0.0001 55.77 { 15.82 (76) 0.0310 (NS)CAC96d 43.91 { 14.42 (76) 49.95 { 17.01 (73) 0.0127 53.95 { 19.14 (76) 0.2324 (NS)

Average 38.37 { 16.47 (637) 47.61 { 16.25 (601) õ0.0001 54.86 { 17.53 (152) 0.5242 (NS)PSA-N-CAM (5A5)

CAC18 24.25 { 15.24 (84) 23.22 { 10.99 (82) 0.7614 (NS)CAC35 40.68 { 18.57 (124) 30.29 { 19.84 (108) õ0.0001CAC82d 36.65 { 18.95 (108) 37.69 { 12.94 (100) 0.2995 (NS) 45.10 { 20.31 (98) 0.0178CAC92d 50.62 { 19.46 (116) 43.76 { 19.56 (109) 0.0149 40.47 { 22.50 (115) 0.3282 (NS)CAC99d 29.70 { 14.16 (105) 54.64 { 18.07 (107) õ0.0001 62.18 { 16.70 (103) 0.0016

Average 37.30 { 19.65 (537) 38.66 { 19.99 (506) 0.1779 (NS) 52.04 { 23.17 (316) 0.0004

Note. Numbers in parentheses show number of growth cones measured. NS, not significant; P ú 0.02.a TMN neurons were grown in media supplemented with 10 ng/ml BDNF.b TMN neurons were grown in media supplemented with 10 ng/ml BDNF, NT-3, and NGF.c Comparison of TMN neurons grown in media supplemented with BDNF with those grown in media supplemented with all three

neurotrophic factors.d Antibody supplied by Dr. Urs Rutishauser (see Methods).

DRG (Honig and Kueter, 1995). In particular, levels of Our cultures of TMN neurons were grown in defined me-dium supplemented only with BDNF, although in vitroexpression of L1, axonin-1, BEN, and N-cadherin were

lower on trigeminal muscle afferents, relative to cuta- TMN neurons can be supported by, and express high-af-finity receptors for, both NT-3 and BDNF (Davies et al.,neous afferents, than in DRG. This suggested the possibil-

ity that our cultures lacked something required by muscle 1986; Hohn et al., 1990; Williams et al., 1995). In contrast,DRG neurons were grown in a rich mixture containingafferents for full expression of their repertoire of CAMs.



106 Scott et al.

FIG. 2. Histograms summarizing the relative levels of expression of CAMs on DM-TG and TMN neurons. Results from three to sixexperiments (see Table 1) are pooled for each histogram. Arrows indicate the mean level of expression.

muscle conditioned medium, horse serum, and chick em- 1995). Although TMN neurons do not express high-affin-ity trk receptors for NGF (Williams et al., 1995) and arebryo extract supplemented with NGF (Honig and Kueter,

1995). It is well established that NGF can significantly not supported by NGF in vitro (Davies et al., 1987), theycould potentially respond to NGF via low-affinity recep-influence the phenotype of sensory neurons, independent

of cell survival (Lindsay and Harmar, 1989; Lewin et al., tors (cf. Seilheimer and Schachner, 1987; Johnson et al.,1988; Itoh et al., 1995). Thus, the possibility existed that1992; Mandelzys and Cooper, 1992; Bevan and Winter,




inclusion of NGF in the media could alter the profile of mean { SD (n Å 21)] on L1 than neurites emanating fromTMN explants [26.3 { 9.8 mm/hr (n Å 20); P õ 0.01], asCAMs expressed on muscle sensory neurons.

To determine whether expression of CAMs on TMN neu- assayed with time lapse microscopy. In contrast, TMN neu-rites grew much faster [70.3 { 9.8 mm/hr (nÅ 18)] than DM-rons might be increased by addition of these other trophic

factors, we grew TMN neurons in a mixture of NGF, NT- TG neurites [43.9 { 12.4 mm/hr (n Å 18); P õ 0.001] onlaminin.3, and BDNF. Addition of NGF and NT-3 did not alter ex-

pression of axonin-1, BEN, L1, N-cadherin, or N-CAM on We also measured the length of neurites of individualdissociated neurons at a single time point in three separateTMN neurons. In contrast, PSA-N-CAM was more abun-

dant on TMN neurons supplemented with all three neuro- experiments. Surprisingly, there was no difference in theaverage length of the longest neurite of TMN and DM-TGtrophic factors (Table 1), raising the intriguing possibility

that sialation of N-CAM on sensory neurons may be regu- neurons on L1 (Table 2), despite the fact that DM-TG neu-rites appeared to grow faster on L1. TMN neurites were,lated by factors produced by the target.however, significantly longer on laminin than on L1 andwere also significantly longer than DM-TG neurites on lam-

Effects of Age on CAM Expression inin. Possible reasons for the apparent discrepancy betweengrowth rate and neurite length on L1 are discussed below.Many of the CAMs studied here are developmentally reg-

The morphology of trigeminal sensory neurons grown onulated, with their level of expression changing during em-L1 differed from their morphology on laminin. In general,bryogenesis. The DRG neurons studied by Honig and Kuetergrowth cones of both TMN and DM-TG were broader andwere from E7 (St. 30–31) embryos, whereas trigeminal neu-more lamellar on L1 than on laminin, as previously reportedrons studied here were from E10 (St. 36) embryos. Thus, thefor chick retinal neurons (Payne et al., 1992). Neuritesapparent differences in CAM expression between DRG andbranched more extensively on L1 than on laminin, as evi-trigeminal neurons may simply reflect differences in the agedenced by the appearance of both explants and dissociatedof embryos studied. To test this possibility, we examinedneurons. Growth cones from explants on L1 took more tor-expression of axonin-1 on trigeminal sensory neurons fromtuous paths and interacted more frequently with adjacentE61

2–712 (St. 29–31) embryos. We selected axonin-1 for these

growth cones than on laminin. This was particularly strik-studies since the difference in its expression on DM-TG and ing for TMN explants, in which neurites tended to growTMN in E10 embryos was obvious without quantification. out individually or as small straight fascicles on laminin,As with older embryos, TMN neurons of E7 embryos did but which formed a dense meshwork on L1 (Fig. 3). Similar,not express detectable levels of axonin-1, whereas DM-TG but less dramatic, differences were also seen for DM-TGneurons did. Moreover, addition of embryo extract and explants. In addition, both types of neurons branched sig-horse serum to cultures did not induce the expression of nificantly more extensively when grown as dissociated neu-axonin-1 on E7 TMN neurons. We cannot rule out the possi- rons on L1 than on laminin (Table 2).bility that differences in expression of other CAMs betweenDRG and trigeminal sensory neurons may be related to age;the sparse outgrowth from young TMN explants precluded

DISCUSSIONa thorough analysis of CAM expression at this age.

CAM Expression on Trigeminal Sensory NeuronsEffects of CAM Expression on Neurite Growth and

Previously we have shown that growth cones of avianMorphologytrigeminal cutaneous and muscle afferents respond differ-ently to several potential targets in vitro (Woodbury andTo test whether the observed difference in CAM expres-

sion are sufficient to have functional consequences, we Scott, 1995). The goal of the present experiments was tocompare the expression of surface CAMs on these neurons.compared the growth of DM-TG and TMN neurites on L1,

which is a potent substrate for neurite growth (Lagenaur We found that at E10 cutaneous neurons from the dor-somedial portion of the avian trigeminal ganglion (DM-TG)and Lemmon, 1987). We predicted that DM-TG neurites

would adhere better and grow faster than TMN neurons express higher levels of axonin-1, BEN, L1, and N-cadherinthan do muscle afferents from the TMN. The latter expresson L1, because DM-TG growth cones express significantly

more of the CAMs that bind L1 [L1 (Lemmon et al., 1989) relatively higher levels of N-CAM, whereas polysialated-N-CAM (PSA-N-CAM) is similar on the two populations.and axonin-1 (Kuhn et al., 1991; see also Sakurai et al.,

1994)] than do TMN growth cones. For comparison, we also A tacit assumption in these studies is that the characteris-tics of growth cones observed in the simplified in vitro envi-assayed the growth of both populations of neurons on lami-

nin, since laminin is frequently used as a substrate for sen- ronment mirror their in vivo characteristics. This appearsto be the case, for at least some of the CAMs studied here.sory neurite growth in vitro.

As expected, neurites emanating from DM-TG explants For example the pattern of axonin-1 expression that we ob-served on DM-TG and TMN neurons mimics that reportedgrew slightly, but significantly, faster [35.7 { 12.6 mm/hr,



108 Scott et al.

in vivo (Halfter et al., 1994; see also Sakurai et al., 1994;but see Karagogeos et al., 1991 for mammalian TAG-1).Many of the other molecules we tested are expressed ontrigeminal sensory neurons in vivo, as described below, al-though the relative levels of expression on TMN and DM-TG neurons have not been compared. Based on what isknown, however, the pattern of CAMs described here mostlikely represents real differences between DM-TG andTMN neurons.

The reported spatial and temporal distributions of theadhesion molecules that we studied suggest that they couldbe involved in patterning trigeminal sensory axon out-growth during embryonic development. For example, manyof these molecules, such as BEN (Chedotal et al., 1995),axonin-1 (Sakurai et al., 1994), and N-cadherin (Redies et al.,1992) are expressed only transiently on trigeminal sensoryneurons, being down-regulated when axon growth is com-plete. L1 (NgCAM) is widely distributed on axons in thebrainstem of embryonic chicks (Sakurai et al., 1994) andthus could provide a permissive substrate for outgrowth oftrigeminal axons that express either L1 or axonin-1 (Lem-mon et al., 1989; Kuhn et al., 1991). Axonin-1 (SC2) has amore limited distribution, being expressed on specific nervefiber bundles, such as trigeminal axons (Halfter et al., 1994;Sakurai et al., 1994). The restricted expression of axonin-1suggests it may be involved in the selective fasciculationof trigeminal axons, allowing them to segregate from otherL1-expressing axons in the brainstem. N-cadherin, whichmediates homophilic binding, may also be important in se-lective fasciculation and patterning of trigeminal axons. Itis expressed on some, but not all, developing trigeminalsensory axons in the chick, with trigeminal axons that ex-press N-cadherin being spatially segregated from those thatdo not (Redies et al., 1992). Much less is known about therole of these molecules on TMN neurons in vivo. WhereasBEN, which also binds homophilically (El-Deeb et al., 1992;DeBernardo and Chang, 1995), is expressed on developingTMN neurons, it may not be important in their selectivefasciculation, as it is also expressed on numerous other ax-ons that TMN axons appear to ignore (Chedotal et al., 1995).

Several of the molecules we examined, such as N-CAMand N-cadherin, are also expressed in developing targetmuscle (Fredette et al., 1993) and/or skin (Jiang and Chuong,1992) at the time that innervation is being established.Whereas these adhesion molecules have clear effects on thedevelopment of the target per se (Jiang and Chuong, 1992;Fredette et al., 1993), their role in patterning sensory in-nervation has not been examined. It seems unlikely, basedon our results of neurons growing on L1, that the differencein expression of any single CAM on DM-TG and TMNneurons is sufficient to discriminate between these twopopulations of neurons with respect to their pathway ortarget choice. However, experiments that perturb the func-tion of individual or several CAMs (e.g., Landmesser et al.,

TA

BL

E2

Mor

phol

ogy

ofD

isso

ciat

edN

euro

ns

onL

amin

inan

dL

1

TM

Non

lam

inin

TM

Non

L1

DM

-TG

onla

min

inD

M-T

Gon

L1

Exp

erim

ent

Bra

nch

Bra

nch

Bra

nch

Bra

nch

no.

Len

gth

(mm

)po

ints

Len

gth

(mm

)po

ints

Len

gth

(mm

)po

ints

Len

gth

(mm

)po

ints

SMC

738

8.30{

113.

74(1

7)**

, ††3.

65{

2.87

253.

24{

74.6

4(1

8)8.

68{

6.92

*25

5.07{

101.

79(2

7)3.

11{

2.76

234.

02{

104.

09(2

2)8.

27{

5.63

**SM

C8

519.

26{

209.

29(1

7)**

, ††2.

35{

2.00

214.

59{

90.2

4(1

6)4.

50{

3.39

*23

7.93{

117.

51(1

9)2.

79{

1.72

198.

33{

95.5

3(2

4)4.

88{

1.87

**K

F130

3.27{

151.

22(1

2)*, †

2.58{

1.88

183.

67{

79.1

6(1

5)9.

27{

8.33

*19

2.66{

106.

74(1

0)2.

20{

2.15

152.

02{

63.0

7(1

5)10

.80{

7.24

**

*0.

01°

Põ

0.05

;**

Põ

0.00

1fo

rco

mpa

riso

ns

betw

een

one

type

ofn

euro

non

diff

eren

tsu

bstr

ates

.†

0.01°

Põ

0.05

;††

Põ

0.00

1fo

rco

mpa

riso

ns

betw

een

TM

Nan

dD

M-T

Gon

the

sam

esu

bstr

ate.

1988; Stoeckli and Landmesser, 1995) will be required to




FIG. 3. Video images of explants of E10 TMN neurons grown on laminin (A) and L1 (B). Note the extensive branching and large flattenedgrowth cones (arrows) of neurons growing on L1. Bar, 120 mm.

elucidate the specific role of these CAMs in embryonic de- medium containing BDNF, whereas DRG neurons weregrown in a mixture of muscle conditioned medium, embryovelopment.extract, horse serum, and NGF. We found, however, that addi-tion of NGF and NT-3 to TMN cultures had little effect on

Comparison with CAM Expression on DRG expression of most CAMs. Moreover, small amounts of horseNeurons serum and chick embryo extract were also insufficient to turn

on expression of axonin-1 in TMN neurons, the only CAMThe pattern of expression of CAMs on trigeminal sensorystudied under these conditions. Nevertheless, we cannot ruleneurons differs markedly from that reported recently for cuta-out the possibility that maximal expression of CAMs requiresneous and muscle afferents in DRG of E7 embryos, with TMNsome other factor present in the rich, but undefined, mediumneurons expressing relatively less of most CAMs than theirin which DRG neurons were grown, but which is lacking incounterparts in DRG. One possible explanation for this dis-

crepancy is that TMN neurons were usually grown in defined defined medium.



110 Scott et al.

Interestingly, addition of NGF and NT-3 appeared to in- ferences in the behavior of growth cones emanating fromexplants compared to those of dissociated neurons (Honigcrease expression of the polysialated form of N-CAM on

TMN neurons, without altering expression of N-CAM per and Burden, 1993). Nevertheless, these experiments to-gether suggest that the differences in L1 and axonin-1 ex-se, as if these additional neurotrophic factors promote siala-

tion of N-CAM. Although contact with the target, which pression on DM-TG and TMN growth cones are not suffi-cient to have major effects on neurite behavior, at leastpresumably is the normal source of such trophic factors,

has been shown to regulate expression of highly sialated N- in the assays we used. Similarly, there is little correlationbetween adhesion and growth rate of retinal growth conesCAM, the effects here are opposite to those normally seen

in vivo, where sialation often decreases after target contact on different substrates (Lemmon et al., 1992).In contrast, TMN neurons grew much faster and extended(Tosney et al., 1986; Bruses et al., 1995).

Another possible explanation for the discrepancies be- longer neurites on laminin than DM-TG neurons. This con-trasts with our previous study where DM-TG and TMNtween CAM expression on DRG and trigeminal sensory

neurons is that the two studies used embryos of different grew at approximately the same rate on laminin (Woodburyand Scott, 1995); in these earlier studies DM-TG neuritesages. Each study represents a snap shot of growth cones at

a single age, yet expression of many CAMs changes during grew more rapidly on laminin than in the present studies.In our previous studies dishes were coated with both polyor-development. Thus, it is possible that the apparent discrep-

ancies between the two studies arose in part because we nithine and laminin, whereas in the present studies theywere coated with nitrocellulose and laminin. We do notsampled CAMs at different stages in their life histories.

This cannot be the full explanation, however, because TMN know the final concentration of laminin that attached tothe dishes in either case, but it is unlikely to be responsibleneurons did not express axonin-1 in vitro at E7; nor do these

neurons express axonin-1 at any stage in vivo (Halfter et for the observed differences in growth rates (Beuttner andPittman, 1991). A more likely explanation is that polyor-al., 1994), whereas DRG muscle afferents do (Honig and

Kueter, 1995; but see Halfter et al., 1994). Thus, it appears nithine enhances the growth of DM-TG neurites more thanTMN neurites, and therefore DM-TG neurites do not ex-that TMN neurons may not be equivalent to muscle affer-

ents in DRG, despite their common neural crest origin. press their full growth potential on laminin attached tonitrocellulose.Since we and others (e.g., Davies et al., 1986; Hohn et al.,

1990) have used TMN as models of muscle afferents, it will These findings call into question the notion that neuronshave an ‘‘intrinsic’’ rate of growth, which is correlated withbe of interest to investigate other aspects of their morpho-

logical, physiological, and biochemical characteristics to the distance that they must grow in vivo and which is main-tained and expressed in vitro (Davies, 1989b). During em-learn the extent to which they are comparable to muscle

afferents in DRG. bryonic development TMN neurites have farther to growthan DM-TG to reach their peripheral targets, yet they growat the same rate as DM-TG neurons on polyornithine/lami-

Neurite Growth on L1 and Laminin nin in vitro (Woodbury and Scott, 1995); this is the substrateused by Davies to demonstrate ‘‘intrinsic’’ rates of growth.Growth rates. Axon extension on L1 is mediated by

both homophilic binding of L1 (Lemmon et al., 1989) and In contrast, TMN neurites extend much more rapidly thanDM-TG on laminin alone and more slowly on L1. Thus,heterophilic binding of axonin-1 (Kuhn et al., 1991; see also

Sakurai et al., 1994). Because DM-TG neurons express while neurites may have an optimum or maximum growthrate, the speed that they extend in vitro appears to dependhigher levels of these ligands than TMN neurons, we pre-

dicted that they would grow more rapidly on L1. Indeed on the substrate, at least for regenerating neurons such asthose studied here (see also Lemmon et al., 1992).neurites emanating from DM-TG explants on L1 grew

slightly, but significantly, faster than those emanating from Morphology. The morphology of DM-TG and TMNoutgrowth also differed between laminin and L1. Out-TMN explants. In contrast, DM-TG and TMN produced

neurites of the same average length when grown as dissoci- growth from explants of DM-TG and TMN were less fascic-ulated on L1 than on laminin. Similar differences have beenated isolated neurons. One possible explanation for this dis-

crepancy is that our time lapse recordings of necessity sam- observed for retinal cells growing on L1 (Lemmon et al.,1992) and for DRG neurons growing on JC7 (BEN) (El-Deebpled the longest neurites to avoid interactions with neigh-

boring growth cones (see also Honig and Burden, 1993). et al., 1992). Since the degree of fasciculation of neuronsdepends on the relative adhesivity of adjacent neurites andThus, in these experiments we actually compared the fast-

est growing TMN neurites with the fastest growing DM-TG the substrate (Rutishauser et al., 1978; Landmesser et al.,1988), it appears that L1 is more adhesive than laminin forneurites. In contrast, in dissociated cultures we randomly

sampled neurons independent of their growth rate. Appar- both types of sensory neurons, as has been demonstrateddirectly for avian retinal neurons (Lemmon et al., 1992). Inently the fastest DM-TG neurites outpaced the fastest TMN

neurites on L1, but the majority of the two populations addition, isolated DM-TG and TMN neurons branchedmore extensively on L1 than on laminin. Neurons tend togrew at similar rates. This may not be the full explanation,

however, as other investigators have observed marked dif- elaborate a greater number of branches on more adhesive




of chick sensory neurons to brain-derived neurotrophic factor. J.substrates (Letourneau, 1975; Bray et al., 1987), again sug-Neurosci. 6, 1897–1904.gesting that L1 is a more adhesive substrate for both types

DeBernardo, A. P., and Chang, S. (1995). Native and recombinantof neurons.DM-GRASP selectively support neurite extension from neuronsthat express GRASP. Dev. Biol. 169, 65–75.

Dodd, J., Morton, S. B., Karagogeos, D., Yamamoto, M., and Jessell,ACKNOWLEDGMENTS T. M. (1988). Spatial regulation of axonal glycoprotein expression

on subsets of embryonic spinal neurons. Neuron. 1, 105–116.El-Deeb, S., Thompson, S. C., and Covault, J. (1992). Characteriza-We thank David H. Adams for his patient assistance with the

tion of a cell surface adhesion molecule expressed by a subset offigures and Dr. Mahendra S. Rao for helpful discussions and com-developing chick neurons. Dev. Biol. 149, 213–227.ments on an earlier version of the manuscript. Antibodies 5A5, 5e,

Fredette, B., Rutishauser, U., and Landmesser, L. (1993). RegulationBEN, 23.4-5 3A10 and EC/8 were obtained from the Developmentaland activity-dependence of N-cadherin, NCAM isoforms, andStudies Hybridoma Bank maintained by the Department of Phar-polysialic acid on chick myotubes during development. J. Cellmacology and Molecular Sciences, Johns Hopkins UniversityBiol. 123, 1867–1888.School of Medicine (Baltimore, MD, 21205) and the Department of

Grumet, M. (1992). Structure, expression, and function of Ng-CAM,Biological Sciences, University of Iowa (Iowa City, IA 52242) undera member of the immunoglobulin superfamily involved in neu-Contract N01-HD-2-3144 from the NICHD. We thank Dr. Urs Rut-ron-neuron and neuron-glia adhesion. J. Neurosci. Res. 31, 1–13.ishauser for his gift of 5A5, 5e, and N-cadherin antibodies, Dr. Carl

Halfter, W., Yip, Y. P. L., and Yip, J. W. (1994). Axonin 1 is expressedLagenaur for L1, and Regeneron Pharmaceuticals, Inc., for BDNFprimarily in subclasses of avian sensory neurons during out-and NT-3. This work was supported by NIH Grant NS 16067 togrowth. Dev. Brain Res. 78, 87–101.S.A.S.

Hamburger, V., and Hamilton, H. L. (1951). A series of normalstages in the development of the chick embryo. J. Morphol. 88,49–92.

REFERENCES Hohn, A., Leibrock, J., Bailey, K., and Barde, Y.-A. (1990). Identifi-cation and characterization of a novel member of the nervegrowth factor/brain-derived neurotrophic factor family. NatureBevan, S., and Winter, J. (1995). Nerve growth factor (NGF) differen-344, 339–341.tially regulates the chemosensitivity of adult rat cultured sensory

Honig, M. G. (1991). Differential expression of adhesion moleculesneurons. J. Neurosci. 15, 4918–4926.on subpopulations of sensory neuron growth cones. Soc. Neu-Bray, D., Bunge, M. B., and Chapman, K. (1987). Geometry of iso-rosci. Abst. 17, 209.lated sensory neurons in culture. Effects of embryonic age and

Honig, M. G., and Burden, S. M. (1993). Growth cones respond inculture substratum. Exp. Cell Res. 168, 127–137.diverse ways upon encountering neurites in cultures of chickBruses, J. L., Oka, S., and Rutishauser, U. (1995). NCAM-associateddorsal root ganglia. Dev. Biol. 156, 454–472.polysialic acid on ciliary ganglion neurons is regulated by polysia-

Honig, M. G., and Kueter, J. (1995). The expression of cell adhesionlyltransferase levels and interaction with muscle. J. Neurosci. 15,molecules on the growth cones of chick cutaneous and muscle8310–8319.sensory neurons. Dev. Biol. 167, 563–583.Buettner, H. M., and Pittman, R. N. (1991). Quantitative effects of

Itoh, K., Brackenbury, R., and Akeson, R. A. (1995). Induction oflaminin concentration on neurite outgrowth in vitro. Dev. Biol.L1 mRNA in PC12 cells by NGF is modulated by cell-cell contact145, 266–276.and does not require the high-affinity NGF receptor. J. Neurosci.Burns, F. R., von Kannen, S., Guy, L., Raper, J. A., Kamholz, J., and15, 2504–2512.Chang, S. (1991). DM-GRASP, a novel immunoglobulin super-

Jiang, T.-X., and Chuong, C.-M. (1992). Mechanism of skin morpho-family axonal surface protein that supports neurite extension.genesis. I. Analyses with antibodies to adhesion molecules tenas-Neuron. 7, 209–220.cin, N-CAM, and integrin. Dev. Biol. 150, 82–98.Chedotal, A., Pourquie, O., and Sotelo, C. (1995). Initial tract forma-

Johnson, E. M., Jr., Taniuchi, M., and DiStefano, P. S. (1988). Expres-tion in the brain of the chick embryo: Selective expression of thesion and possible function of nerve growth factor receptors onBEN/SC1/DM-GRASP cell adhesion molecule. Eur. J. Neurosci.Schwann cells. Trends Neurosci. 11, 299–304.7, 198–212.

Karagogeos, D., Morton, S. B., Casano, F., Dodd, J., and Jessell,D’Amico-Martel, A., and Noden, D. M. (1983). Contributions ofT. M. (1991). Developmental expression of the axonal glycopro-placodal and neural crest cells to avian cranial peripheral ganglia.tein TAG-1: Differential regulation by central and peripheralAm. J. Anat. 166, 445–468.neurons in vitro. Development. 112, 51–67.Davies, A. M. (1989a). Neurotrophic factor bioassay using dissoci-

Kuhn, T. B., Stoeckli, E. T., Condrau, M. A., Rathjen, F. G., andated neurons. In ‘‘Nerve Growth Factors’’ (R. A. Rush, Ed.), pp.Sonderegger, P. (1991). Neurite outgrowth on immobilized axo-95–109. Wiley, New York.nin-1 is mediated by a heterophilic interaction with L1(G4). J.Davies, A. M. (1989b). Intrinsic differences in the growth rate ofCell Biol. 115, 1113–1126.early nerve fibres related to target distance. Nature 337, 553–

Lagenaur, C., and Lemmon, V. (1987). An L1-like molecule, the555.8D9 antigen, is a potent substrate for neurite extension. Proc.Davies, A. M., Lumsden, A. G. S., and Rohrer, H. (1987). NeuralNatl. Acad. Sci., USA 84, 7753–7757.crest-derived proprioceptive neurons express nerve growth factor

Landmesser, L. (1994). Axonal outgrowth and pathfinding. In ‘‘Prog-receptors but are not supported by nerve growth factor in culture.ress in Brain Research’’ (F. J. Seil, Ed.), Vol. 103, pp. 67–73. Else-Neuroscience. 20, 37–46.

Davies, A. M., Thoenen, H., and Barde, Y.-A. (1986). The response vier, New York.



112 Scott et al.

Landmesser, L., Dahm, L., Schultz, K., and Rutishauser, U. (1988). Rutishauser, U., Gall, W. E., and Edelman, G. M. (1978). Adhesionamong neural cells of the chick embryo. IV. Role of the cellDistinct roles for adhesion molecules during innervation of em-surface molecule CAM in the formation of neurite bundles inbryonic chick muscle. Dev. Biol. 130, 645–670.cultures of spinal ganglia. J. Cell Biol. 79, 382–393.Lemmon, V., Burden, S. M., Payne, H. R., Elmslie, G. J., and Hlavin,

Sakurai, T., Shiga, T., Shirai, T., Tanaka, H., and Grumet, M. (1994).M. L. (1992). Neurite growth on different substrates: permissiveBiochemical characterization and immunolocalization of SC2versus instructive influences and the role of adhesive strength.protein: SC2 is indistinguishable from the cell adhesion moleculeJ. Neurosci. 12, 818–826.axonin-1. Dev. Brain Res. 83, 99–108.Lemmon, V., Farr, K. L., and Lagenaur, C. (1989). L1-mediated axon

Scott, S. A. (Ed.) (1992) ‘‘Sensory Neurons: Development, Diversityoutgrowth occurs via a homophilic binding mechanism. Neuronand Plasticity.’’ Oxford Univ. Press, New York.2, 1597–1603.

Scott, S. A., Chau, C. A., and Randazzo, K. F. (1994). CutaneousLemmon, V., and McLoon, S. C. (1986). The appearance of an L1-and muscle afferents: Differential expression of adhesion mole-like molecule in the chick primary visual pathway. J. Neurosci.cules in vitro. Soc. Neurosci. Abst. 20, 656.6, 2987–2994.

Seilheimer, B., and Schachner, M. (1987). Regulation of neural cellLetourneau, P. C. (1975). Cell-to-substratum adhesion and guidanceadhesion molecule expression on cultured mouse Schwann cellsof axonal elongation. Dev. Biol. 44, 92–101.by nerve growth factor. EMBO J. 6, 1611–1616.Lewin, G. R., Ritter, A. M., and Mendell, L. M. (1992). On the

Silos-Santiago, I., Jeng, B., and Snider, W. D. (1995). Sensory affer-role of nerve growth factor in the development of myelinatedents show appropriate somatotopy at the earliest stage of projec-

nociceptors. J. Neurosci. 12, 1896–1905.tion to dorsal horn. NeuroReport. 6, 861–865.

Lindsay, R. M., and Harmar, A. J. (1989). Nerve growth factor regu- Stoeckli, E. T., and Landmesser, L. T. (1995). Axonin-1, Nr-CAM,lates expression of neuropeptide genes in adult sensory neurons. and Ng-CAM play different roles in the in vivo guidance of chickNature 337, 362–364. commissural neurons. Neuron 14, 1165–1179.

Mandelzys, A., and Cooper, E. (1992). Effects of ganglionic satellite Takeichi, M. (1988). The cadherins: Cell-cell adhesion moleculescells and NGF on the expression of nicotinic acetylcholine cur- controlling animal morphogenesis. Development 102, 639–655.rents by rat sensory neurons. J. Neurophysiol. 67, 1213–1221. Tanaka, H., and Obata, K. (1984). Developmental changes in unique

Mirnics, K., and Koerber, H. R. (1995a). Prenatal development of cell surface antigens of chick embryo spinal motoneurons andrat primary afferent fibers: I. Peripheral projections. J. Comp. ganglion cells. Dev. Biol. 106, 26–37.Neurol. 355, 589–600. Tosney, K. W., Watanabe, M., Landmesser, L., and Rutishauser, U.

Mirnics, K., and Koerber, H. R. (1995b). Prenatal development of (1986). The distribution of NCAM in the chick hindlimb duringrat primary afferent fibers: II. Central projections. J. Comp. Neu- axon outgrowth and synaptogenesis. Dev. Biol. 114, 437–452.rol. 355, 601–614. Williams, R., Backstrom, A., Kullander, K., Hallbook, F., and Eben-

Narayanan, C. H., and Narayanan, Y. (1978). Determination of the dal, T. (1995). Developmentally regulated expression of mRNAembryonic origin of the mesencephalic nucleus of the trigeminal for neurotrophin high-affinity (trk) receptors within chick tri-nerve in birds. J. Embryol. Exp. Morphol. 43, 85–105. geminal sensory neurons. Eur. J. Neurosci. 7, 116–128.

Woodbury, C. J., and Scott, S. A. (1995). Cutaneous and musclePayne, H. R., Burden, S., and Lemmon, V. (1992). Modulation ofafferents: Interactions with potential targets in vitro. J. Neurosci.growth cone morphology by substrate-bound adhesion mole-15, 3936–3951.cules. Cell Motil. Cytoskeleton 21, 65–73.

Zuellig, R. A., Rader, C., Schroeder, A., Kalousek, M. B., Von BohlenPourquie, O., Coltey, M., Thomas, J.-L., and Le Douarin, N. M.Und Halback, F., Osterwalder, T., Inan, C., Stoeckli, E. T., Af-(1990). A widely distributed antigen developmentally regulatedfolter, H.-U., Fritz, A., Hafen, E., and Sonderegger, P. (1992). Thein the nervous system. Development 109, 743–752.axonally secreted cell adhesion molecule, axonin-1. PrimaryRathjen, F. G., Wolff, J. M., Frank, R., Bonhoeffer, F., and Rutis-structure, immunoglobulin-like and fibronectin-type-III-like do-hauser, U. (1987). Membrane glycoproteins involved in neuritemains and glycosylphosphatidylinositol anchorage. Eur. J. Bio-fasciculation. J. Cell Biol. 104, 343–353.chem. 204, 453–463.Redies, C., Inuzuka, H., and Takeichi, M. (1992). Restricted expres-

sion of N- and R-cadherin on neurites of the developing chicken Received for publication April 1, 1996Accepted June 21, 1996CNS. J. Neurosci. 12, 3525–3534.



differential expression of cell adhesion molecules on ... · differential expression of cell...

Documents