Growth cone guidance by substrate-bound laminin pathways is correlated with neuron-to-pathway adhesivity

DEVELOPMENTAL

BIOLOGY

Growth JAMES

126,29-39

(1988)

Cone Guidance by Substrate-Bound Correlated with neuron-to-Pathway

A. HAMMARBACK,*~~

JAMES

*Department

B. MCCARTHY,?

SALLY

L.

PALM,?

LEO

Laminin Pathways Adhesivity T. FURCHT,~

AND

PAUL

Is

C. LETOURNEAU*

of Cell Biology and hburoanatmy, and ~Department of Laboratory Medicine and Patholcgg, University of Minnesota, 321 Church. Street SE., Minneapolis, Minnesota 55455

Substrate-bound Iaminin pathways prepared by the method of Hammarback et al. [J. A. Hammarback, S. L. Palm, L. T. Furcht, and P. C. Letourneau (1985). J. Neurosci. Res. 13, 213-Zi!O] guided peripheral nervous system neurjtes (dissociated dorsal root ganglia and sympathetic ganglia) and central nervous system neurites (dissociated spinal cord and brain). Guidance of individual growth cones by 7- to lo-pm-wide laminin pathways was observed using time-lapse video microscopy. Fibronectin pathways, produced by the method used for laminin pathways, did not guide neurites. The guidance effect of laminin pathways was quantified and found to correlate with the concentration of laminin initially applied to the substratum. The concentration of laminin initially applied to the substratum also correlated with increased adhesivity of dorsal root ganglia (DRG) neurons to laminin constituting the pathways relative to uv-irradiated laminin that borders the pathways. The guidance effect of laminin pathways was blocked by anti-lamlnin antibodies or by laminin but not by anti-flbronectin antibodies. This study demonstrates that guidance of DRG neurites by laminin occurs at the growth cone in a manner consistent with the hypothesis of guidance by differential neuronto-substratum adhesivity. Q 1988 Academic Press, hc. INTRODUCTION

sivity in axonal guidance by naturally occurring materials has not been well studied. This paper describes a semiquantitative measure of cell guidance elicited by laminin and fibronectin molecules which may provide guidance in v&o. Using a recently devised method for creating ‘7- to lo-pm pathways of undamaged molecules (Hammarbaek et al., 1985), we have measured the guidance of axons by narrow pathways of substratumbound molecules in the presence of an alternative substratum that supports neurite extension. The pathways are made without introducing coincident changes in substratum shape that would make identification of guidance effecters more difficult. Time-Iapse video microscopy is used to visualize guidance effects of laminin pathways on neurite extension. The responsiveness to laminin is investigated by determining whether specific antibodies block the guidance effect. In addition, the cell-type specificity of guidance by laminin and fibronectin pathways is examined by culturing both central and peripheral neurons and several non-nerve cell lines on the pathways. Finally, the relationship between guidance and neuron-substratum adhesivity is analyzed by quantitation of neuron adhesivity to undamaged laminin substratum relative to that to the alternative substratum of uv-irradiated laminin.

Morphogenesis of neuronal connections is precise (e.g., Ho and Goodman, 1982; Lance-Jones and Landmesser, 1981) and may depend on adhesive guidance of extending axons by extracellufar cues (Berlot and Goodman, 1984; Gaudy and Bentley, 1986; Nardi, 1983; Schlosshauer et al., 1984; Silver and Rutishauser, 1984). Several extracellular matrix molecules are suspected of having a role in guidance of vertebrate axons. Among these, the large glycoproteins laminin (-800 kDa) and ~bronectin (~400 kDa) are of particular interest. Both laminin and fibronectin are found along many of the pathways characteristically taken by extending axons, including basement membranes (Palm and Furcht, 1983; Sanes et ok, 1978; Taylor and Roberts, 1983; Wan et a& 1984; Wartiovaara et ah, 1980). A laminin and fibroneetin receptor is found in neuronal cultures (Bozyczko and Horwitz, 1986). The initiation and growth of neurites in vitro is promoted by substratum-bound laminin and fibronectin (Rogers et al., 1983). In addition, pathways of substratum-bound laminin guide neurite outgrowth (Smallheiser et al., 1984; Hammarback et al., 1985).

Neurites follow pathways of greatest cell-substratum adhesivity when cultured on artificial substrata of patterned adhesivity (Letourneau, 1975). The role of adhe-

MATERIALS *To whom correspondence Group, Worcester Foundation MA 01545.

should be addressed at Cell Biology for Experimental Biology, Shrewsbury,

Cell culture. sympathetic 29

AND

METHODS

Chick dorsal root ganglia (DRG) and ganglia from 8- to IO-day embryos were 0012-1606/88 Copyright 0 All rights

$3.00

1988 by Academic Press, Inc. of reproduction in any form resewed.

30

DEVELOPMENTALBIOLOGY VOLUME126,1988

dissected and dissociated as described previously (Rogers et al, 1983). Cells plated on patterned substrata were cultured in serum-free F14 medium (GIBCO) supplemented with 5 lg/ml insulin, 10 rig/ml B nerve growth factor (NGF), 5 rig/ml sodium selenite, 100 @g/ml human transferrin (Sigma), 2 rnn/l glutamine, and 1X antibiotic-antimycotic mixture (GIBCO). Cultures were maintained in 5% COZ and 95% room air in a humidified atmosphere. Spinal cords from &day embryos were dissociated, as described for DRG, and plated in serum-free F14 medium with the supplements listed above, except that no NGF was added. Brain from lo-day embryos was dissected into three crude regions: pre-tectal, tectum, and medulla. Brain tissue was dissociated as described above and cultured in serum-free F14 medium with the supplements listed above, except that no NGF was added. Mouse fibrosarcoma (Mixed Met) cells, originally provided by Dr. I. J. Fidler, University of Texas, M. D. Anderson Hospital, Houston, Texas, were cultured in vitro in Dulbecco’s modified Eagle’s medium (DMEM) and passaged twice weekly. RN22F Schwannoma cells were originally provided by Dr. John Sheppard, Dwight Institute of Genetics, University of Minnesota. The RN22F cell line is derived from a chemically induced rat Sehwannoma, RN22 (Pfeiffer and Wechsler, 1972). RN22F cells were maintained in DMEM containing 10% heat-inactivated horse serum and passaged at 5- to &day intervals. Approximately 20,000 cells were plated onto ultravioletlight patterned coverslips resting in 35mm plastic petri dishes (Falcon). S~~s~ra~u ~~~~ar~~~o~. The method for preparing laminin and fibronectin pathways is modified slightly from that described by Hammarback et ~2. (1985). Purified laminin or fibronectin was applied to nitric acidcleaned coverslips that had been extensively rinsed with distilled water and sterilized with dry heat. Laminin from the mouse EHS tumor and human plasma fibronectin were isolated as previously described (Rogers et ah, 1983). Preliminary experiments have also been done using the 600-kDa thrombin fragment of laminin, produced and purified as previously described (Palm et aL, 1985). One hundred microIiters of laminin or fibronectin at 100 pglml in 0.05 iM carbonate buffer (pH 9.6) containing 0.2 mg/ml sodium azide was applied to one side of a coverslip. After 1 hr at room temperature, the protein-treated coverslips were rinsed with distilled water and air-dried. Substratum pathways were made by placing electron microscope grids directly on protein-treated air-dried coverslips before irradiating the coverslips with ultraviolet light for 2 hr with a bactericidal lamp (Westinghouse Sterilamp 782L-30) at a distance of 40 cm. After irradiation, the electron mi-

croscope grids were removed, uncovering the invisible grid patterns of unirradiated protein, Immunochemistry. Affinity-purified rabbit anti-laminin or anti-fibronectin, described in Rogers et al. (1983), were used to detect substratum patterns of unirradiated laminin or fibronectin, respectively. Cells on uv-light patterned substrates were fixed with 4% paraformaldehyde and 0.2% glutaraldehyde in phosphatebuffered saline (PBS, pH 7.4) for 20 min at room temperature. The coverslips were then incubated with 1 mg/ml sodium borohydride in PBS for 15 min and incubated with 100 ~1 of 0.23 yg/ml rabbit anti-laminin or anti-fibronectin followed by a rhodamine-conjugated goat anti-rabbit secondary antibody (Cappel) as described previously (Letourneau, 1981). In experiments to study the effects of antibodies on patterned neurite outgrowth, coverslips with the uv-patterned laminin substrata were treated with anti-laminin or anti-fibronectin antibody at 23 pug/ml in PBS for 1 hr, rinsed, and then treated for 1 hr with a 11100 dilution of rhodamine-conjugated goat anti-rabbit secondary antibody, The unbound secondary antibody was removed by rinsing the coverslips and then cells were plated. Immunofluorescent staining with the A2B5 antibody, kindly provided by Dr. Sherry Rogers, University of New Mexico, was used to estimate the percentage of substratum-bound cells of neuronal origin (Eisenbarth et ab, 1979; Schnitzer and Schachner, 1982). In adhesivity assays where the A2B5 antibody was used, the antibody was applied to living cells for 10 min, and the cells were subjected to high shear (see below), fixed with 4% paraformaldehyde in PBS, rinsed, and then incubated with rhodamine-conjugated goat anti-mouse secondary antibody (Cappei). Quantitation of guidance eflects. Guidance of axons by laminin and fibronectin pathways was quantified by measuring the percentage of a neurite’s length found on a protein pathway. DRG cells were cultured for 1 day on substrata containing 7- to lo-pm-wide laminin or fibronectin pathways. Cells were then fixed with 4% paraformaldehyde in PBS and processed for immunofluorescence to detect the protein pathways. Total neurite length was taken to be the distance from the proximal point of neurite contact with an unirradiated pathway to the neurite tip or the most distal identifiable point along the neurite. The length of neurite in contact with the protein pathway, past the initial contact, was measured and then divided by the total neurite length. Only neurites having a contact with pathways were measured. Laminin binding. C3H]laminin (Jentoft and Dearborn, 1979) was used to determine the loss of unirradiated versus uv-irradiated laminin from coverslips during

HAMMARBACKETAL.

31

Adhesive Laminin Pathways

cell culture. Coverslips were coated with [3H]laminin as described for unlabeled laminin. Some coverslips were then uv-irradiated for 1 hr at a distance of 40 cm from the lamp. All coverslips were rinsed with 1.5 ml of serum-free medium, the rinse was saved for counting by liquid scintillation (Beckman LS 6800). The counts in this rinse are reported as “counts in medium” after 0 days in culture (Table 5). Triplicate uv-irradiated and unirradiated coverslips were evaluated for [3H]laminin bound after 1 day of culture in 1.5 ml serum-free medium with or without 20,000 DRG cells/dish. These coverslips were rinsed with serum-free medium, crushed, and boiled 10 min in glass scintillation vials containing 1 ml of 1% sodium dodecyl sulfate followed by liquid scintillation counting. The 24-hr culture medium and the medium used to rinse the coverslips were also counted. Adhesion assays. Cell attachment to uv-irradiated versus unirradiated laminin and fibronectin substrata was measured by determination of resistance to liquid shearing forces. Halves of lamininand fibronectintreated coverslips were uv-irradiated to create sharp borders between unirradiated versus uv-irradiated adsorbed protein. Individual coverslips were placed in 35-mm petri dishes with 40,000 dissociated dorsal root ganglion cells and cultured for 3 hr in medium containing 10% fetal calf serum or 10% fibronectin-depleted fetal calf serum (see specific experiments below). Fibronectin-depleted fetal calf serum was prepared by sequential passages over an agarose-gelatin column (Rogers et ah, 1983). Some coverslips were then subjected to low shear by normal processing for immunofluorescence. Other cell cultures were subjected to high shear by placing the dishes on a rotatory shaker for 1 min at 200 rpm, followed by fixation and processing for immunofluorescence. Cell counts were made in paired fields of view that were one microscope field (40x objective) removed from the border between unirradiated and uv-irradiated protein. Ten paired fields per coverslip were counted, two or three individually treated coverslips per shear condition (see specific experiments below). In experiments where A2B5 immunofluorescence was used to identify neuronal cells bound to the substratum, all the cells along both sides of the laminin-uv-laminin border were counted. Gel electrophwesis. Laminin, 0.66 ml of 100 pug/ml in Voller’s carbonate buffer, pH 9.6, was applied to each of six acid-cleaned glass slides (1.5 X 3 in.) for overnight. The slides were rinsed with dHBO and air-dried 15 min, and three slides were uv-irradiated 2 hr at 40 cm from the light source. Laminin was removed from the uv-irradiated and unirradiated coverslips by scraping with a rubber policeman and using a small volume of 1% so-

dium dodecyl sulfate (SDS). Samples were separated on a 3 to 8% SDS-polyacrylamide gradient gel with a 2% stacking gel using the gel formulations of Blattler et al. (1972) and the buffer system of Laemmli (1970). Gels were stained with Coomassie brilliant blue G-250 and destained with 10% acetic acid. Time-lapse video microscopy. Observations of living cells on substratum-bound laminin pathways were made on coverslips mounted over holes drilled in 50-mm petri dishes. A mixture of bee’s wax, white petrolatum, and paraffin (1:l:l) was melted and used to seal the coverslips into the dishes. The coverslips were then treated with laminin and patterned with ultraviolet light. The protein pathways were made visible by lightly staining the protected laminin with 2.3 pg/ml anti-laminin for 30 set followed by rinsing and incubation with l/100 rhodamine-conjugated goat anti-rabbit antibody for 30 sec. DRG cells were cultured overnight on the patterned substrata in F12 buffered with 10 mM Hepes and supplemented as described above. The cultures were then placed on a Zeiss IM microscope stage, heated to 37°C with an air-stream incubator (Nicholson Precision Instruments). A SIT camera (Model 65; Dage-MTI, Inc.) and a time-lapse video recorder (Panasonic NV-8030) were used to record the motility of the extending neurites. The lowest light intensity that permitted image formation was used to record neurite outgrowth. A heat filter and green filter were placed in the illuminating light path. While the video-tape was replayed at 30 frames/set, 35-mm pictures were taken from the monitor, using Tech-Pan film with 0.25- to l.O-see exposures. RESULTS

Time-lapse video recordings show that individual growth cones can precisely follow laminin pathways (Fig. 1). To visualize pathways during time-lapse recordings, pathways were briefly exposed to low concentrations of rabbit anti-laminin and then rhodamineconjugated goat anti-rabbit antibodies before plating DRG cells. Longer exposure of laminin pathways to higher concentrations of these antibodies will block the guidance effect of laminin pathways (Table 1). Growth cones on the 7- to lo-pm-wide laminin pathways are narrow with few observations of the broad lamellae that are seen on some growth cones on unpatterned laminin substrata. Filopodia of growth cones on laminin pathways extend over and contact the uv-irradiated substratum but are retracted (Fig. 1). Quantitation of guidance was accomplished by observing the position of extended neurites and growth cones relative to the position of unirradiated laminin

32

D~VELOFME~AL

BIOLOGY

VOLUME I%,1988

and fibronectin pathways. As measures of guidance, the percentage of neurite length remaining on pathways distal to the point of initial pathway contact was calculated and the number of neurites with growth cones on pathways was recorded. Measurements were made only on neurites and growth cones that encountered a pathway as judged by their position on or across a pathway. The percentage of neurite length on pathways is a reliable indicator of guidance if neurite to pathway adhesion is sufficient because the forming neurite remains attached along the pathway taken by its growth cone. If neurite to pathway adhesion is low, then guided growth cones should accumulate on pathways even though the percentage of neurite length on pathways is low. The absence of growth cone accumulation on pathways under conditions where the percentage of neurite length on pathways is low indicates that growth cones are not guided (or immobilized) by laminin or fibronectin pathways (Tables 1 and 2). Quantitation of guidance shows that the guidance effect of laminin pathways is increased by applying a higher concentration of laminin to the substratum that is used to create pathways by uv-irradiation. For a given concentration of laminin, increasing the exposure of unprotected laminin to uv-irradiation also increases the guidance effect (Table 2). The percentage of a neurite’s length found on a laminin pathway was 92% when 50 Kg/ml of laminin was used and unprotected laminin received 2 hr of irradiation. Neurite guidance on laminin pathways is blocked by adsorbing additional laminin onto coverslips bearing laminin pathways (Table 1). Anti-laminin antibodies block neurite outgrowth on laminin patterns, if the patterns are incubated with 23 pg/ml rabbit anti-laminin antibody followed by rhodamine-conjugated goat anti-rabbit antibody before plating the cells (Table 1). Laminin pathways elicit patterned neurite outgrowth from cells of the peripheral and central nervous system.

FIG. 1. Individual growth cones are guided by laminin pathways. Photographs were taken during the replay of a time-lapse video recording of a neurite extending on a laminin pathway. A sequence of four photographs, in chronological order from top to bottom, is displayed, The numbers in each photograph indicate the date, hour, and minute when the living neurite was observed (the seconds are blurred

because several frames from the video-recording were averaged to produce each photograph). Dashed lines have been drawn in to indicate the borders of the narrow laminin pathways as detected by immuno~uorescence (see Materials and ~eth~s). The growth cone is able to extend filopodia over the uv-irradiated laminin but neurite extension proceeds directly along the laminin pathway. In the second photograph, the growth cone momentarily deviates toward an alternate laminin pathway but proceeds straight ahead. Another momentary deviation from the laminin pathway is seen in the bottom photograph. In this case, the neurite also continued straight along the pathway and eventually contacted the cell soma seen at the far right of the bottom photograph. Another neurite, slightly out of focus, is seen aligned with the vertical laminin pathway in the bottom photograph, The field of view in the bottom photograph is immediately to the right of the field of view in the top three photographs (each field of view is 98 pm wide).

33

Adhesive LamininPathways

HAPIIMARBACK ETAL.

TABLE

1

ANTIBODY EFFECFS ONNEURITE GUIDANCE Percentage of neurite length on pathway

Growth cones on pathway/growth cones off pathway

8 4

95 2 4 (40) 3 + 3 (40)

21/o 2129

141* 24 83_tl7 100 t 11

85 r 7 (20) 22&9(20) 4 + 1 (20)

14/z 3/8 o/4

Mean Substratum treatment

Substratum pathway Experiment 1 Laminin Fibronectin Experiment 2 Laminin Laminin Laminin

-

neurite length (pm) 73k 56t

Anti-fibronectin Anti-laminin Laminin

Note. Substratum pathways were made by applying 100 pl of laminin (50 pg/ml) or fibronectin (100 pg/ml) to glass coverslips that were then used in creating narrow unirradiated pathways of these glycoproteins. Dissociated DRG cells were cultured for 1 day on the pathways. The guidance effect of substratum pathways was assayed by determining the percentage of neurite length on the pathways (CC+ SE, N) and the number of these neurites having growth cones on pathways. Neurite length was measured from the most proximal point of intersection between neurite and pathway to the most distal identifiable point of the neurite. The total number of growth cones is less than the number of neurites measured because growth cones in contact with other cells could not always be identified. In Experiment 2, substratum pathways were treated with rabbit anti-laminin or anti-fibronectin antibodies and then goat anti-rabbit antibodies before DRG cells were plated (see Materials and Methods).

Cells from sympathetic ganglia, brain, and spinal cord align with laminin pathways (Figs. 2 and 3). Time-lapse observations of these cells were not made, therefore the relationship of neurite patterning to non-nerve cell patterning, if any, is not known. Laminin pathways also align the initial attachment and spreading of the Schwannoma cell line, RN22F, and a mouse fibrosarcoma line, Mix Met (Fig. 4). Pathways of fibronectin that were prepared using uv-irradiation, as for the laminin pathways, did not elicit guided neurite outgrowth from peripheral or eentral neurons. The pathways of unirradiated fibronectin can clearly be identified using anti-fibronectin antibodies (Figs. 2 and 3). DRG and sympathetic neurites ex-

tend on both the uv-irradiated fibroneetin and the unirradiated fibronectin pathways, and they do not seem to follow the pathways. Although incapable of guiding neurite elongation, these pathways of unirradiated fibronectin do have an effect on the initial orientation of the mouse fibrosarcoma cells (Fig. 4). Alignment of the fibrosarcoma cells is striking within a few hours of plating the cells on fibronectin pathways, but the alignment deteriorates with time in culture and with increased cell density. The RN22F Schwannoma cells show no obvious orientation to pathways of unirradiated fibronectin (Fig. 4). Thus, a difference between uv-irradiated and protected fibronectin is seen by some cell types. Because cells use different regions of fibro-

TABLE 2

EFFECTS OFLAMININCONCENTRATION ANDUV-IRRADIATION ONNEURITE GUIDANCE Substratum pathway Laminin Laminin Laminin Laminin Laminin Laminin

(20 (30 (40 (50 (50 (50

pg/ml) pg/ml) ag/ml) rg/ml) Fg/ml) rg/ml)

Note. Substratum pathways in creating narrow unirradiated effect of substratum pathways theee neurites having growth a Exposure of non-pathway

Minutes of exposure to uv light* 120 120 120 120 60 30

Neurite length (pm) 7oi71+ 102 + 211 + 152 f 84 f

2 7 18 27 25 10

Growth cones on pathway/~owth cones OS pathway

Percentage of neurite length on pathway

O/16 o/21 8/9 12/2 20/l 14/6

5k5(22) 0 (31) 46 It 9 (26) 92 zk 5 (22) 83 It 8 (23) 49 rfr8 (29)

were made by applying 100 pl of the designated concentration of laminin to glass coverslips that were later used pathways of these glyeoproteins. Dissociated DRG cells were cultured for 1 day on the pathways. The guidance was assayed by determining the percentage of neurite length on the pathways (z r SE, N) and the number of cones on pathways. See Materials and Methods for further details. areas of the substratum to uv-irradiation.

LAM

FN

HAMMARBACKETAL.

Adhesive

Laminin

Pathways

35

LAM

za

FIG. 4. Response of non-nerve cells to fibronectin and laminin pathways. Schwannoma (RN22F) and Mixed Met (MIX MET) cells on laminin and fibronectin pathways. Simultaneous phase-contrast and epifluorescent microscopy was used to see the relationship between cells and immunofluorescently labeled pathways. Mixed Met cells were oriented by laminin pathways. Schwannoma cells bound and spread preferentially on 30-pm laminin pathways but their orientation varied. The Schwannoma cells did not attach well to the lo-pm pathways that the Mixed Met cells are photographed on. The Mixed Met cells were the only cells that were obviously oriented by fibronectin pathways. The alignment of Mixed Met cells on fibronectin pathways deteriorates with increased cell density and with increased time in culture.

nectin for attachment (McCarthy et ah, 1986), this result suggests that the sensitivity of different regions of fibronectin to uv-irradiation may differ. Greater adhesion of growth cones and filopodia to the unirradiated laminin paths may be responsible for the guidance of neurite outgrowth. In experiments where extending neurites had the opportunity to extend on more than one substratum, the preferred substratum had the greatest adhesivity (Letourneau, 1975). Greater adhesion of total DRG cells to unirradiated versus uv-

irradiated laminin is indicated by the resistance of DRG cells to displacement from these substrata by shearing forces produced on a rotatory shaker (Table 3). Immunofluorescent detection of the A2B5 antigen, an antigen specific for cells of neuronal origin (Eisenbarth et al., 1979; Schnitzer and Schachner, 1982), was used to estimate the percentage of attached DRG cells that are neurons (Fig. 5). There is not a statistically significant reduction in the percentage of total DRG cells that are A2B5 positive when shearing forces are increased. The

FIGS. 2 AND 3. Response of peripheral (Fig. 2) and central (Fig. 3) nervous system cells to fibronectin (FN) and laminin (LAM) pathways. Dorsal root ganglia (DRG), sympathetic ganglia (SYMP), spinal cord (CORD), and brain cells (BRAIN) all exhibited patterned growth on laminin pathways but none of these cell types demonstrated obvious patterning on fibronectin pathways. The relationship between cells and immunofluorescently labeled pathways is seen using simultaneous phase-contrast and epifluorescent microscopy. The patterning of CORD cells on LAM is evident using phase-contrast microscopy alone. BRAIN cell patterning on LAM appeared to be caused by increased cell density on the laminin pathways consisting of rounded cells associated with fibroblast-like cells; extensive neurite outgrowth from BRAIN cells on LAM did not occur. BRAIN and CORD cells on FN extended short neurites that did not appear to align with FN pathways.

36

DEVELOPMENTAL

BIOLOGY

VOLUME

126,1988

TABLE 3 ADHESIVITY OF DRG CELLS TO UNIRRADIATED VERSUS UV-IRRADIATED GLYCOPROTEINS

Shear force applied to cells Low High Low High

Cell substratum Laminin uv-Laminin Laminin uv-Laminin Fibronectin uv-Fibronectin Fibronectin uv-Fibronectin

Cells remaining after applying shear forces 82? 7 60 f 11 80 f 18 29 ic 20 70f 6 109 + 5 7Ok 0 73 f 14

Percentage change in number of cells bound after uvirradiation -27*

4

-66f

9

56+

6

04 ? 14

Note. Halves of laminin- and fibronectin-treated coverslips were uv-irradiated to create sharp borders between unirradiated versus uv-irradiated adsorbed protein. DRG cells were plated on these coverslips for 3 hr in fibronectin-depleted fetal calf serum and then subjected to low shear by processing them for immunofluorescence or subjected to high shear by placing cultures on a rotatory shaker at 200 rpm for 1 min before processing them for immunofluorescence. Ten sets of paired 40X objective lens fields of view, one field on unirradiated substrata and the other adjacent field on uv-irradiated substrata, were counted on each coverslip (z k SE, n = 2). See Materials and Methods for details of adhesivity assay.

percentage A2B5 positive cells went from 34 + 5 to 22 + 6% on laminin and from 33 + 3 to 29 k 1% on uv-irradiated laminin when the shearing forces were increased. Therefore, major changes in total DRG cells bound are indicative of a change in neurons bound in this adhesion assay. As the amount of laminin applied to coverslips is increased, the relative difference in adhesion to unirradiated versus uv-irradiated laminin is increased (Table 4). Recall from Table 2 that over this same range of laminin concentrations, the magnitude of the guidance effect of laminin pathways is increased. Thus, increased guidance by laminin pathways correlates with a greater relative difference in neuronal adhesivity to the protected versus uv-irradiated laminin. In contrast to the results with laminin, more DRG cells were found on uv-irradiated fibronectin than on unirradiated fibronectin under either the low or high shear conditions (Table 3). This finding coincides with the absence of neurite guidance along pathways of fibronectin protected from uv-irradiation (Table 1). Note that under the low shear condition approximately 56% more cells were found on uv-irradiated fibronectin while under the high shear condition approximately equal numbers of cells remained bound to uv-irradiated and unirradiated fibronectin. Therefore, we can not rule

FIG. 5. Staining of DRG cells with A2B5 antibody to detect cells of neuronal origin. Immunofluorescent detection of the A2B5 antigen was performed on cells plated for adhesivity assays to determine the percentage of cells bound having neuronal origins. The same field of view is seen by phase-contrast microscopy (top) and epifluorescent microscopy (bottom). The antibody specifically stained round phasebright cells that had begun to extend cell processes but did not stain flattened phase-dark cells.

out the possibility that DRG cells or a subpopulation of DRG cells attach more rapidly but with lower adhesivity to uv-irradiated fibronectin than to unirradiated fibronectin. TABLE RELATIVE DRG CELL UV-IRRADIATED LAMININ APPLIED TO SUBSTRATUM

Laminin applied to coverslip 25 @g/ml 50 pg/ml 100 pg/ml

4

ADHESIVITY OF UNIRRADIATED AND VERSUS CONCENTRATION OF LAMININ

Percentage change in number of cells bound after uv-irradiation +16 + 12 -47 f 3 -61 + 12

Note. The designated concentration (100 ~1) of laminin was applied to glass coverslips used in adhesivity assays. DRG cells were plated for 3 hr on the designated substrata in fetal calf serum and then subjected to high shear by placing cultures on rotatory shaker at 200 rpm for 1 min before processing them for immunofluorescence. Ten sets of paired 40X objective lens fields of view were counted on each coverslip as described under Materials and Methods (z + SE, N = 3).

HAMMARBACK

Adhesive

ET AL.

The patterning of cells observed in Fig. 1 might be due to a greater loss of uv-irradiated laminin from the substratum in the presence of cells and culture media over 24 hr. However, the results in Table 5 show that the amount of unirradiated [‘Hllaminin bound to coverslips after 24 hr in culture is not significantly different from the amount of uv-irradiated [3H]laminin that remains bound (Table 5). When cpm released into the culture medium are measured, a statistically significant difference (Student’s t test, P < 0.05) in release of [3H]laminin from unirradiated versus uv-irradiated r3H]laminin is detected in 0 day cultures in the presence of cells and in 1 day cultures in both the presence and absence of cells. After 1 day of culture without cells, about 13% of the counts bound at Day 0 are released from unirradiated PH]laminin while 19% of the total cpm bound are released from uv-irradiated [3H]laminin, After 1 day of culture with cells, 11% of cpm bound were released from unirradiated [3H]laminin as were 15% of the cpm of uv-irradiated [3H]laminin. It seems unlikely that this difference in cpm released to the medium can account for the guidance effect or increased adhesivity of unirradiated laminin. The time-lapse video observations of neurites extending on laminin pathways indicate that guidance effects of laminin pathways are not simply the result of the detachment of neurites that are not coincidently aligned with pathways of unirradiated laminin, In addition, the presence of DRG cells and therefore the motile activities of DRG

BINDING

OF [aHjL~~~~~~

no cells

Laminin treatment

No cells No cells Cells

No uv uv No uv

Cells

or

cells

No cells No cells Cells Cells

uv

No uv uv No uv uv

TABLE 5 TO COVERSLIPS

IN CULTURE

Days in culture

Counts in medium

Counts bound to eoverslip

0 0 0

379+ 69 492+ 20 308% 23 433* 7 1638 k 93 2388f 270 1471 + 65 2096 k137

12,567 f 1,063

0

1 1 1 1

-

10,807 + 10,349 + 10,157 f 10,178f

405 763 894 364

Note. [3H]laminin (50 &g/ml; 19,340 + 1724 counts; 2 r SE, N = 3) applied to coverslips as described under Materials and Methods. Culture medium (1.5 ml) was briefly applied to 0 day cultures and then removed and counted. Another 1.5 ml of medium was applied to the same coverslips and they were cultured for 1 day. Then the medium was removed, and coverslips were rinsed with another 1.5 ml of rinse medium. The “Counts in medium” for 1 day cultures includes the medium that was on the coverstips overnight and the 1.5 ml of rinse medium. The coverslips were then placed in 1 ml of SDS in a liquid scintillation vial, crushed, and boiled for 10 min. Liquid scintillation counting was used to estimated [3H]laminin in the medium and bound to coverslips (z + SE, N = 3). was

Lam&in

37

Pathways

cells do not detectably increase the cpm of either unirradiated or uv-irradiated [3H]laminin released to the culture medium. The appearance of substratum material in the culture medium does not necessarily indicate that cells will be easily displaced from the substratum. For instance, even though a significant amount of unirradiated and uv-irradiated [3H]laminin is released from the substratum during 1 day in culture, increasing shearing forces do not cause a significant change in the number of cells bound to unirradiated laminin although about 50% of cells bound to uv-irradiated laminin are removed by high shear conditions (Table 3). We do not know how surface-bound laminin is altered by uv-irradiation; preliminary results from slab gel electrophoresis suggest that some of the substratum laminin may be cross-linked by uv-irradiation (data not shown). DISCUSSION

Neurites will not extend in liquid culture medium (Harrison, 1910); they require a growth matrix or substratum. Neurites extending on a low adhesivity substratum frequently adhere to the substratum at the growth cone only. These observations suggest that the direction and rate of neurite extension could be determined by adhesive contacts between growth cones and their environment. These adhesive contacts may act to promote ~ytoskeletal modi~cations within the growth cone that pull the growth cone in the direction of the adhesion (see Letourneau, 1983). The evidence is strong for regulation of neurite extension rates and direction by artificial materials in vitro (Letourneau, 1975). In viva studies on guidance effects of extracellular materials and the possible role of cell to extracellular matrix adhesivity are difficult to interpret. In general, in &JO studies must rely on morphology to assess adhesivity. In vivo studies have identified materials likely to be contacted by growth cones. Such studies reveal basement membrane components and integral membrane glycoproteins of other cells as extracellular terrain for growth cones (Palm and Furcht, 1983; Sanes et al, 1975; Taylor and Roberts, 1983; Wan et al, 1984; Wartiovaara et al, 1980; Schlosshauer et ab, 1984). Testing the efficacy of these potential guidance factors in vitro complements di~cult in viva experiments. The findings of these studies are consistent with the hypothesis that guidance of neurite outgrowth along these laminin pathways is related to differential neuron-substratum adhesion. Unirradiated laminin has greater adhesivity for A2B5 antigen positive DRG cells than uv-irradiated laminin (Table 3). This difference in adhesivity is correlated with the amount of laminin applied to coverslips (Table 4), and with the effectiveness of laminin pathways in a guidance of neurites

38

DEVELOPMENTAL BIOLOGY

(Table 2). However, the relationship between adhesion to laminin of DRG neuronal cell bodies and adhesion of the growth cone, where neurite guidance occurs, remains to be established. To date, there are no reports that the cell body differs qualitatively or quantitatively in adhesivity from the growth cone. Asymmetric distributions of an integral membrane glycoprotein (Ellis et aZ., 1985a,b) and lectin binding sites (Pfenninger and Maylie-Pfenninger, 1981a,b, 1984) on neurites have been reported and indicate the potential for differences in adhesivity mediated by glycoproteins or other lectin binding molecules. A putative fibronectin and laminin receptor recognized by the CSAT antibody has been localized on the perikarya, neurites, and growth cones of chick peripheral neurons (Bozyczko and Horwitz, 1986). The morphology of growth cones extending on 7- to lo-km laminin paths is consistent with a difference in adhesivity between laminin and uv-irradiated laminin. Growth cone lamellae and filopodia are primarily directed forward along the laminin pathways. The extent of growth cone spreading is found to be correlated with growth cone-substratum adhesivity to artificial substrata (Letourneau, 1979). Effects of ultraviolet light on substratum-adsorbed proteins used in these studies are not well documented in the literature, and chemical studies will be required to determine the critical molecular effects of uv-irradiation on adsorbed laminin that result in the cellular guidance effect illustrated in Figs. 2 and 3. uv-Irradition may directly modify the neuron binding sites of laminin or may produce conformational changes that block the interaction of neurons with these sites. It is also possible that uv-irradiation cross-links laminin at some sites and cleaves it at other sites. The failure of some fibroblastic cells from DRGs to align on the unirradiated laminin pathways (Hammarback et aZ., 1985), the ability of additional laminin or anti-laminin antibodies to block guidance and the failure of fibronectin pathways to guide neurites, while orienting other cell types, indicate that guidance is a result of specific cell responses and not a result of nonspecific effects of uvirradiated proteins on cell motility and shape. The adhesivity of DRG cells to adsorbed fibronectin is not lowered by uv-irradiation as judged by adhesivity assay. The immunoreactivity of fibronectin is sensitive to uv-irradiation but is, qualitatively, more resistant to uv-irradiation than laminin immunoreactivity. If fibronectin pathways are produced by overnight uv-irradiation, the immunoreactivity of the uv-irradiated fibronectin is qualitatively lower than it is after 2 hr of irradiation, but the fibronectin pathways still fail to guide neurites. Therefore, the DRG cell binding site on fibronectin may be insensitive to ultraviolet light even

VOLUME 126.1988

though fibronectin immunoreactivity is sensitive to ultraviolet light. Cell-type specific guidance effects were not found in a study using artificial substrata (Harris, 1973). Previous failures to identify cell-type specific effects may be related to the fact that guidance effects of substratumbound materials had been tested primarily with substrata composed of artificial materials such as metals and synthetic polycations. More work is needed to determine the basis of the cell-type specific effects of laminin pathways. The use of uv-light patterned substrata has demonstrated quantifiable guidance effects of a naturally occurring molecule, laminin, of potential significance for development and regeneration of the nervous system. The relative ease in measuring the guidance effects of pathways of uv-sensitive proteins should facilitate study of the cellular and molecular mechanisms of chemically based axonal guidance. This work was supported by grants from the National Institutes of Health (HD 19950 and HD 1’7192), the Spinal Cord Society, and a Minnesota Medical Foundation, the Muscular Dystrophy Association, and a National Science Foundation to P.C.L., and a grant from the Leukemia Task Force and Grants CA29995 and CA21436 from the National Institutes of Health to L.T.F. REFERENCES BERLOT, J., and GOODMAN, C. (1984). Guidance of peripheral pioneer neurons in the grasshopper: Adhesive hierarchy of epithelial and neuronal surfaces. Science 223,493-496. BLATTLER, P., GARNER, F., VAN SLYKE, K., and BRADLEY, A. (1972). Quantitative electrophoresis in polyacrylamide gels. J. Chromatogr. 64,147-155.

BOZYCZKO,D., and HORWITZ, A. F. (1986). The participation of a putative cell surface receptor for laminin and fibronectin in peripheral neurite extension. J. Neurosci. 6, 1241-1251. CAUDY, M., and BENTLEY, D. (1986). Pioneer growth cone morphologies reveal proximal increases in substrate affinity within leg segments of grasshopper embryos. J. Neurosci. 6,364-379. EISENBARTH, G. S., WALSH, F. S., and NIRENBERG, M. (1979). Monoclonal antibody to a plasma membrane antigen of neurons. Proc Natl.

Acad.

Sci. USA 76,4913-4917.

ELLIS, L., WALLIS, I., ABREN, E., and PFENNINGER, K. H. (1985a). Nerve growth cones isolated from fetal rat brain. IV. Preparation of a membrane subfraction and identification of a membrane glycoprotein expressed on sprouting neurons. J. Cell Biol. 101,1977-1989. ELLIS, L., WALLIS, I., ABREN, E., and PFENNINGER, K. H. (1985b). Immunolocalization of a neuronal growth-dependent membrane glycoprotein. J. Cell Biol 101,1990-1998. HAMMARBACK, J. A., PALM, S. L., FURCHT, L. T., and LETOURNEAU, P. C. (1985). Guidance of neurite outgrowth by pathways of substratum-adsorbed laminin. J. Neurosoi. Res. 13,213-220. HARRIS, A. K. (1973). Behavior of cultured cells on substrata of variable adhesiveness. Exp. Cell Res. 77,285-297. HARRISON, R. G. (1910). The outgrowth of the nerve fiber as a mode of protoplasmic movement. J. &p. Zool. 9,787-848. Ho, R. K., and GOODMAN, C. S. (1982). Peripheral pathways are pio-

HAMMARBACK ET AL.

Adhesive

neered by an array of central and peripheral neurones in grasshopper embryos. Nature (London) 297,404-406. JENTOFT, N., and DEARBORN, D. G. (1979). Labeling of proteins by reductive methylation using sodium cyanoborohydride. J. Biol. Chem. 254,4359-4365. LAEMMLI, U. K. (1970). Cleavage of structural proteins during assembly of the head of bacteriophage T4. Nature (London) 227,680-685. LANCE-JONES, C., and LANDMESSER, L. (1981). Pathway selection by chick lumbosacral motoneurons during normal development. Proc. R. Sot. London B. 214,1-18. LETOURNEAU, P. C. (1975). Cell-to-substratum adhesion and guidance of axonal elongation. Dev. BioL 44,92-101. LETOURNEAU, P. C. (1979). Cell-substratum adhesion of neurite growth cones, and its role in neurite elongation. Exp. Cell Res. 124, 127-138. LETOURNEAU, P. C. (1981). Immunocytochemical evidence for colocalization in neurite growth cones of actin and myosin and their relationship to cell-substratum adhesions. Dev. BioL 85,113-122. LETOURNEAU, P. C. (1983). Axonal growth and guidance. Trends Neurosci 6,451-455. MCCARTHY, J. B., HAGEN, S. T., and FURCHT, L. T. (1986). Human fibronectin contains distinct adhesion- and motility-promoting domains for metastatic melanoma cells. J. Cell BioL 102, 179-188. NARDI, J. B. (1983). Neuronal pathfinding in developing wings of the moth Manduca sexta. Dev. BioL 95,163-174. PALM, S. L., and FURCHT, L. T. (1983). Production of laminin and fibronectin by schwannoma cells: Cell-protein interactions in vitro and protein localization in peripheral nerve in vivo. J. Cell BioL 96, 1218-1226. PALM, S. L., MCCARTHY, J. B., and FURCHT, L. T. (1985). Alternative model for the internal structure of laminin. Biochemistry 24, 7753-7760. PFEIFFER, S. E., and WECHSLER, W. (1972). Biochemically differentiated neoplastic clone of Schwann cells. Proc. NatL Acad Sci. USA 69,2885-2889.

PFENNINGER, K. H., and MAYLIE-PFENNINGER, M. F. (1981a). Lectin labeling of sprouting neurons. I. Regional distribution of surface glycoproteins. J. Cell BioL 89, 536-546. PFENNINGER, K. H., and MAYLIE-PFENNINGER, M. F. (1981b). Lectin

Laminin

Pathways

39

labeling of sprouting neurons. II. Relative movement and appearance of glycoconjugates during plasmalemmal expansion. J. Cell BioL 89,574-589.

PFENNINGER, K. H., and MAYLIE-PFENNINGER, M. F. (1984). Lectin labeling of sprouting neurons. III. Type specific glycoconjugates on growth cones of different origin. Dev. BioL 106,97-108.

ROGERS, S. L., LETOURNEAU, P. C., PALM,S. L., MCCARTHY, J., and FURCHT, L. T. (1983). Neurite extension by peripheral and central nervous system neurons in response to substratum-bound fibronectin and laminin. Dev BioL S&212-220. SANES, J. R., MARSHALL, L. M., and MCMAHAN, U. J. (1978). Reinnervation of muscle fiber basal lamina after removal of myofibers. J. Cell BioL 78,176-198.

SCHLOSSHAUER,B., SCHWARTZ, U., and RUTISHAUSER, U. (1984). Topological distribution of different forms of neural cell adhesion molecule in the developing chick visual system. Nature 310,141-143. SCHNITZER, J., and SCHACHNER, M. (1982). Cell type specificity of a neural cell surface antigen recognized by the monoclonal antibody A2B5. Cell Tissue Res. 224,625-636. SILVER, J., and RUTISHAUSER, U. (1984). Guidance of optic axons in vivo by a preformed adhesive pathway on neuroepithelial endfeet. Dev. BioL 106,485-499. SMALLHEISER, N. R., CRAIN, S. M., and REID, L. M. (1984). Laminin as a substrate for retinal axons in vitro. Dev. Brain Res. 12, 136-140. TAYLOR, J. S. H., and ROBERTS, A. (1983). The early development of the primary sensory neurones in an amphibian embryo. J. EmbryoL Exp.

MerphoL

75,49-66.

TIMPL, R., ROHDE, H., ROBEY, P. G., RENNARD, S. I., FOIDART, J.-M., and MARTIN, G. R. (1979). Laminin, a glycoprotein from basement membranes. J. BioL Chem 254.9933-9937. WAN, Y., WV, T., CHUNG, A., and DAMJANOV, I. (1984). Monoclonal antibodies to laminin reveal the heterogeneity of basement membranes in the developing and adult mouse tissues. J. Cell BioL 98, 971-979. WARTIOVAARA, J., LEIVO, I., and VAHERI, A. (1980). Matrix glyeoproteins in early mouse development and in differentiation of teratocarcinoma cells. In “The Cell Surface; Mediator of Developmental Processes” (S. Subtelny and N. K. Wessells, Eds.), pp. 305-324. Academic Press, New York.