Attachment of neural crest cells to endogenous extracellular matrices

THE ANATOMICAL RECORD 219:275-285 (1987)

Attachment of Neural Crest Cells to Endogenous Ext racellular Matrices

PHILIP R. BRAUER AND ROGER R. MARKWALD Department ofAnatomy 6 Cellular Biology, Medical College of Wisconsin,

Milwaukee, WZ 53226

ABSTRACT Newly emerging neural crest (NC) cells will enter either the lateral pathway under the surface ectoderm or the vental pathway along the neural tube depending on’the axial level (Pratt et al.: Deu. Biol., 44:298-305, 1975; Thiery et al.: Deu. BioL, 93:324-343, 1982; Newgreen et al.: Cell Tissue Res., 221.521-549, 1982; LeDouarin et al.: In: The Role ofExtracellular Matrix in Deuelopment. Alan R. Liss, Inc., New York, pp. 373-398, 1984; Brauer et al.: Anat. Rec., 211.57-68, 1985). A number of studies have shown a correlation between the type of extracellular matrix (ECM) associated with adjacent tissues (e.g., ectoderm, neural tube, and mesoderm) and the initial pathway taken by NC cells. Our working hypothesis is that the direction of NC cell migration (ventral vs. lateral pathway) depends on the composi- tion of the ECM associated with the surface ectoderm and its ability to support NC cell attachment. In this study, we tested this hypothesis by isolating endogenous ECM associated with the ectoderm of each region and examining the ability of each endogenous ECM to support cranial and trunk NC cell attachment in vitro. Results .indicated that both cranial and trunk NC cells preferentially attached to cranial ectodermal ECM as compared to trunk ectodermal ECM. The differences in NC cell attachment were not due to a preferential adsorption of cranial ectodermal ECM onto the ECM-conditioned plastic substrate over trunk ectodermal since approximately equal amounts of ECM bound to the plastic. These results supported the hypothesis and provide evidence that endogenous ectodermal ECM may be one factor potentially responsible for directing the NC cells along a ventral or a lateral pathway.

Transplantation and chimaeric studies have shown that the characteristic migration patterns of neural crest (NC) cells at different axial levels are imposed by the environment (Noden, 1978a,b; Weston et al., 1978; Le- Douarin et al., 1984). It has been suggested that extracellular matrix (ECM) abundant in areas of NC migration is a primary factor in “specifying” particular NC migratory patterns Weston et al., 1978). Attempts to explore the role of ECM in directing NC cell migration have relied on in vitro experiments using ECM molecules from sources other than potential NC cell pathways (i.e., different species or tissue sites; Newgreen et al., 1982; Rovasio et al., 1983; Erickson and Turley, 1983). However, ECM components can exhibit different structural and functional characteristics depending on their conformation and tissue source. For instance, cellular and plasma fibronectin, although only slightly different in structure (Iwanaga et al., 1978; Crouch et al., 1978; Yamada and Kennedy, 19791, exhibit differences in their ability to bind hyaluronate (Laterra and Culp, 1982), in their ability to promote attachment of transformed cell lines (Hynes et al., 1978), and in the ability of their cell-binding domains to inhibit cell attachment to collagen (Yamada and Kennedy, 1984). Therefore, slight structural differences in ECM components can alter the functional capacities of these ECM components

in vitro. Conformational and functional changes can also occur through associations with other ECM components (Jilek and Hormann, 1979; Johansson and Hook, 1980). Therefore, any in vitro model system for studying ECM interaction with NC cells should ultimately seek to include endogenous sources of ECM and allow multi- ple ECM associations to occur in order to more accu- rately represent in vivo events.

Generally in avian species, cranial NC cells enter the lateral pathway found along the basal surface of the ectoderm and do not enter the underlying mesoderm or ventral pathway between the neural tube and mesoderm (Pratt et al., 1975; Brauer et al., 1985). By contrast, the majority of the initial NC cell population enters the ventral pathway in the trunk axial region he., cell-free space between the somite and the neural tube; Teillet and LeDouarin, 1970; Thiery et al., 1982; Newgreen et al., 1982; LeDouarin et al., 1984; Bronner-Fraser, 1986). The translocation of NC cells is likely to occur both via active translocation (i.e., cells migrate by undergoing repeated attachment to and detachment from a substra-

Received October 20, 1986; accepted April 17, 1987. Philip R. Brauer is now at Department of Biological Chemistry and

Structure, University of Health Sciences, The Chicago Medical School, North Chicago, IL 60064.

0 1987 ALAN R. LISS, INC.

276 P.R. BRAUER AND R.R. MARKWALD

tum) and passive translocation (i.e., cells transported by anchoring to adjacent tissues undergoing differential growth; Noden, 1984). In both cases, cell attachment is essential. Previous studies have suggested that ECM is involved in directing NC cells into the appropriate pathway by mediating NC cell attachment (Greenburg et al., 1981; Newgreen et al., 1982; Erickson and Turley, 1983; Rovasio et al., 1983).

We wanted to determine if the ECM subjacent to the trunk and cranial ectoderm differed in their ability to support NC cell attachment, and if so, whether this difference correlated with the general pattern of NC cell migration which occurs at these levels. If ECM is involved in directing NC cell migration via NC cell attachment, since the majority of NC cells enter the lateral pathway under the surface ectoderm at the cranial axial level and the ventral pathway between the somite and neural tube at the trunk axial level, one would predict that cranial ectodermal ECM and the ECM of the trunk ventral pathway would promote NC cell attachment over that of trunk ectodermal ECM and the ECM of the cranial ventral pathway. In this study, we compared the ability of NC cells from each axial level to attach to isolated, naturally occurring, endogenous ectodermal ECM from similar or reciprocal levels. This study was limited to testing only the ectodermal ECM (i.e., ECM from the lateral pathways of both axial levels) because ECM from the two ventral pathways could not be obtained.

Our results indicated that both cranial and trunk NC cells preferentially attach to cranial ectodermal ECM compared to the trunk ectodermal ECM. The differences in NC cell attachment to ectodermal matrices correlate with the pathway taken by the majority of the NC cell population at these axial levels. These results provide evidence that native endogenous ECM can influence one parameter of NC cell translocation (i.e., attachment) and suggest that ectodermal-associated ECM may provide general directional cues for NC cell migration in both axial levels.

MATERIALS AND METHODS lsolation of Extracellular Matrix

Fertilized white leghorn chicken eggs were incubated at 38°C. Stage 8/9 embryos (Hamburger and Hamilton, 1951) were used to isolate cranial ectodermal ECM and stage 14/15 embryos were used to isolate trunk ectoder- ma1 ECM. Seventy to 80 pieces of ectoderm from the cranial region and trunk region (last seven pairs of somites) were excised in ovo by using tungsten needles and pooled in separate silicone-treated vials containing 200 pl of extraction buffer consisting of calcium-magne- sium-free Dulbecco's buffered salt solution (CMF-DBSS), 5 mM ethylenediamine tetraacetic acid (EDTA), and 0.5 mM phenylmethylsulfonyl fluoride (PMSF). The ectodermal ECM was extracted. Ectoderms and their associated ECM were gently agitated in extraction buffer on a horizontal rotating shaker (80 rpm) for 3 hours a t 4°C; the ectodermal cells were pelleted by centrifugation (40g for 5 minutes); and the supernatant was briefly stored at 4°C. The supernatant fraction contained the extracted ECM (see below). Each cell pellet was washed and placed in 200 pl of CMF-DBSS containing 5 mM EDTA and 0.5 mM PMSF, and the cells were disrupted by sonication. To check for possible cell lysis during the

extraction procedure, the lactic dehydrogenase content of the ECM-containing supernatant relative to the soni- cated ectodermal cells was measured according to Krug et al. (1986).

To determine the extent of ECM extraction, excised ectodermal segments were fixed immediately before or after extraction in 3% glutaraldehyde containing 0.5% cetylpyridinium chloride. Isolated ectodermal segments were then processed by routine methods for transmission electron microscopy and examined on a Hitachi S- 600 transmission electron microscope.

The extracted ECM in the supernatant was precipitated by adding ten volumes of cold (-20°C) acetone for 30 minutes. Acetone was utilized because with its use more protein was precipitated than with either 10% trichloracetic acid or ammonium sulfate precipitation. The mixture was centrifuged at 2,800g for 10 minutes at -20°C. Precipitates were washed four times with water-saturated ether a t -2O"C, once with acetone:ether (1:3, v/v) at -2O"C, and subsequently air dried. The extracted ECM pellets were then rehydrated in 0.5 ml Tyrode's buffer solution (pH 7.31, separated into 100-pl aliquots, and stored in polypropylene tubes at -20°C. Trunk and cranial ectodermal ECM was analyzed for protein content by the method of Peterson (1977) and averaged 5-8 pg of protein per 80 ectodermal segments.

NC Cell Culture Cranial neural folds and tubes (50-60) from stage 9-

11 chick embryos or trunk neural tubes (70-80) from stage 14/15 chick embryos were removed by microdissec- tion. The neural tubes were explanted onto 35-mm tissue culture dishes (ten neural tubes per dish) containing Media 199 supplemented with 1% chick serum, ITS (in- sulin, 5 pg/ml; transferrin, 5pg/ml; selenium, 5 ng/ml; Collaborative Research, Waltham MA), and penicillin, Fungizone, and streptomycin (Gibco 600-5245) and incubated in 5% C02/95% air atmosphere at 37°C. NC cells were allowed to grow out from the neural tube explant, and after 2 days in culture, the neural tubes were removed. The remaining NC cells were washed with serum-free Media 199, resuspended by treatment with 5 mM EDTA in CMF-DBSS for 20 minutes at 4°C and gentle repetitive pipetting, and washed with two changes of serum-free Media 199. Approximately 7 x lo4 to 2 x lo5 NC cells were obtained from the cultured neural tubes. The NC cell concentration was adjusted to 2 x lo4 celldm1 of serum-free Media 199 for the attachment assay. Greater than 90% viability of NC cells was retained based on trypan blue exclusion.

Preparation of Cell Substrates Substrates were prepared by adsorbing the test mate-

rials directly onto the plastic surfaces of 72-microwell plates (10-p1 capacity, A/S Nunc, Denmark). Various concentrations of each test material (see Figs. 2-5) were dissolved in Tyrode's buffer and 10-p1 aliquots were placed in each well and incubated overnight at 4°C. The wells were rinsed by removing the solutions with filter paper and adding two changes of Tyrode's buffer followed by two changes in serum-free Media 199. These substrates were then used immediately in the attachment assay. Test substrates prepared from endogenous ECM included cranial and trunk ectodermal ECM and,

NEURAL CREST ATTACHMENT TO ECM 277

as a physiological control, ECM extracted from stage 14-17 chick embryo hearts by using the same procedure as for ectodermal ECM. Nonendogenous ECM components included human plasma fibronectin (Collaborative Research), bovine serum albumin (BSA; Sigma), and bovine nasal cartilage proteoglycan (kindly provided by Dr. Arnold Caplan, Case Western University, Cleve- land, Ohio) and were used in this study to compare results with those obtained by others (Newgreen et al., 1982; Erickson and Turley, 1983; Rovasio et al., 1983).

To determine whether the percentage of ectodermal ECM components adsorbed to plastic was the same for each region, equal amounts of extracted ECM protein from each region were radioactively labeled by reductive methylation with (3H)KBH4 (Kumarasamy and Symons, 1979) and a known number of cpm were placed into each 10-pl microwell under conditions identical to those in the attachment assays. Subsequently, each well was rinsed four times with Tyrode’s buffer, the rinses were pooled, scintillation fluor was added, and the solution was counted in a Packard Tri-Carb scintillation counter. Additionally, to remove any labeled ECM adsorbed to the plastic, wells were treated with 2% sodium dodecyl- sulfate (SDS)/2-mercaptoethanol followed by 1.0 M sodium hydroxide for 30 minutes at room temperature. Any radioactivity recovered by these treatments was counted.

Cell Attachment Assay Ten-microliter aliquots from a single trunk or cranial

NC cell suspension (approximately 200 cells) were placed into each conditioned microwell and allowed to settle for 10 minutes. This was considered zero time of incubation. NC cells were subsequently incubated for up to 300 minutes on the test substrates at 37°C or 4°C. After incubation, unattached cells were removed by rinsing the entire microwell plate with Tyrode’s buffer four times. Attached cells were fixed with 2% glutaraldehyde and rinsed with buffer; the entire well bottom was pho- tographed; and the remaining cells were counted.

All experiments were repeated two or three times with isolated ECM and NC cells collected at different times. Results did not vary by more than 15%. Data were analyzed by analysis of variance and differences between individual means were determined by two-sided Student’s t-test.

RESULTS Extracted Ectodermal ECM

By using the methods described, small quantities of ECM protein were isolated from each axial ectoderm. After extraction, less than 1% of the lactic dehydrogenase activity was found in the ECM fraction as compared to the ectodermal cell fraction. In addition, electron mi-

croscopic examination of the cell fraction indicated that ectodermal cell membranes were intact after the ECM separation procedure (Fig. lB), and, if placed in culture, the ectodermal cells were viable. Therefore, the extracted material probably did not contain substantial quantities of intracellular material.

Electron microscopic examination of the ectodermal cell fraction further indicated that most of the ECM associated with the excised ectoderm (including the basal lamina) was removed during extraction (Fig. 1). Only sparse patches of basal lamina material remained associated with the cell fraction after the extraction. There- fore, the ECM utilized in the experiments described here contained primarily ectodermal basement membrane (BM) and the ECM immediately subjacent to the BM.

Adsorption of Cranial and Trunk Ectodermal ECM to Plastic Postsynthetic radiolabeling of 0.35 pg of ectodermal

ECM protein by (3H)KBH4 was sufficient to provide greater than 125,000 cpm. Similar quantities of cranial or trunk ectodermal ECM protein bound to the plastic microwells (Table 1). Results indicated that the ECM components were tightly associated with the plastic surfaces and only approximately 80% of the total radioactivity placed in each well could be recovered. Similar observations have been made for plasma fibronectin by Grinnell and Feld (1981) and Haas and Culp (1982).

NC Attachment to Substrates Prepared With Various Concentrations of ECM

The results are shown in Figures 2 and 3. The number of cranial NC cells attached to cranial ectodermal ECM- treated plastic was significantly higher than on untreated plastic. Cranial NC cell attachment to cranial ectodermal ECM was always greater than to trunk ectodermal ECM but was only significant at 5 pg/ml (P < .05). Conversely, trunk NC cells preferentially attached to cranial ectodermal ECM-treated plastic compared to plastic treated with ectodermal ECM from its own axial level at all concentrations (P < .05). Attach- ment of cranial or trunk NC cells to trunk ectodermal ECM-treated plastic was not different from untreated plastic. Since cranial NC cells (somite levels 1-3) are known to eventually populate heart ECM (Kirby et al., 19831, endogenous heart ECM was tested to determine if this ECM supported NC cell attachment in vitro. Results indicated that heart ECM provided significant enhanced attachment at 5pglml for cranial NC cells compared to untreated plastic (Fig. 2).

The results with exogenous ECM components are shown in Figures 2 and 3. Fibronectin significantly enhanced the attachment of cranial and trunk NC cells only at a concentration of 1.0 pg/ml over untreated plastic; increasing concentrations of fibronectin decreased

TABLE 1. ECM adsorbed to plastic’ % removed by

ECM type buffer * SD SDS & NaOH f SD

8.6 * 1.3 Cranial ectodermal ECM Trunk ectodermal ECM 65.7 F 5.6 8.1 f 1.1

P = .61

% removed by

73.6 f 5.3

P = .15

lSee Materials and Methods. N=3.



NC cell attachment. At low concentrations, proteoglycan- and BSA-treated plastic was not different from untreated plastic in supporting NC cell attachment. However, as the proteoglycan and BSA concentration was increased, a decrease in NC cell attachment occurred.

Time Course of Cranial NC Cell Attachment at 37% The time course of cranial NC cell attachment over a

60-minute period at 37°C was examined by using substrates prepared with concentrations of ECM which gave maximum differences from untreated plastic. The results are shown in Figure 4. Cranial NC cell attachment to BSA- and proteoglycan-treated plastic was less than untreated plastic a t all times of incubation. NC cell attachment to cranial ectodermal ECM-treated plastic began to increase compared to untreated plastic after 30 minutes, whereas NC cell attachment to fibronectin and heart ECM-treated plastic did not increase until 60 minutes of incubation. Significant NC cell attachment to trunk ectodermal ECM-treated plastic was greater compared to untreated plastic by 60 minutes (P < .05) but was less than heart ECM-treated plastic and significantly less than cranial ectodermal ECM (P < .01).

Time Course of Cranial NC Cell Attachment at 4OC

NC cell attachment assays at 4°C were identical to that at 37°C except that the incubation time was extended. The purpose of this experimental control was to optimize conditions for observing ECM-ligand interactions in the absence of an interfering metabolic process such as endocytosis and translocation of membrane proteins within the membrane. The pattern of NC cell attachment to the various substrates at 4°C was similar to that obtained at 37°C; however, fewer numbers of NC cells were found attached to the substrates (Fig. 5). Again, NC cell attachment was greater on cranial ectodermal ECM-treated plastic than the other test substrates by 30 minutes. Attachment to heart ECM-treated plastic peaked by 30 minutes but dropped to the level of untreated plastic after 300 minutes of incubation. At- tachment to trunk ectodermal ECM-treated plastic was not significantly different from untreated plastic.

BSA CrM HrM

Abbreviations Bovine serum albumin Cranial ectodermal extracellular matrix Heart extracellular matrix

PFN Plasma fibronectin PG Bovine nasal cartilage proteoglycans PL Untreated plastic TTM Trunk ectodermal extracellular matrix

Fig. 1. Electron micrographs of cranial ectodermal segments before and after EDTA extraction. Surface ectoderm was excised from chick embryos and fixed immediately before (A) or after extraction (B) in 3% glutaraldehyde!0.5% cetylpyridinium chloride and processed for electron microscopy (see Materials and Methods). A A substantial amount of extracellular matrix (ECM) is associated with the excised ectoderm in addition to its basal lamina (arrowheads). ~22 ,560 . B: After extraction with EDTA, most of the ECM and basal lamina material was removed from the ectoderm and only occasional patches of electron- dense material remain (arrowheads). The ectodermal cells did not demonstrate any noticeable discontinuity in their cell membranes. The extracted ECM was then separated from the cells and used in the preparation of the test substrates, as described in the Materials and Methods. ~ 9 , 5 0 0 .

DISCUSSION This study has provided evidence that NC cells pref-

erentially attach to cranial ectodermal ECM (basal lamina and associated ECM) compared to trunk ectodermal ECM (basal lamina and associated ECM) or ECM components derived from nonendogenous sources (i.e., different species or tissues). Previous morphological studies (Weston et al., 1978; Newgreen and Thiery, 1980; Erick- son and Turley, 1983; Brauer et al., 1985) suggest that an ECM characterized by the presence of fibronectin and a diminished sulfated polyanion content would op- timally support NC cell attachment and migration (i.e., just under the surface ectoderm in the cranial region and between the somite and neural tube in the trunk) whereas an ECM characterized by abundant sulfated polyanions would not (i.e., just under the surface ectoderm in the trunk region and the mesodermal intercellular spaces in the cranial region). In addition, studies have shown that when NC cells are reciprocally trans- planted into different axial levels (heterotopic trans- plants), the NC cells follow the NC cell pathway indigenous to the host site (Noden, 1978a,b; LeDouarin, 1980; LeDouarin et al., 1984). Based on these findings one could predict that cranial ectodermal ECM would preferentially support attachment of both cranial and trunk NC cells over that of trunk ectodermal ECM. Results of this study supported this postulate in that both sources of NC cells attached better to cranial ectodermal ECM than to trunk ectodermal ECM. These results were apparently not due to any detectable differences between the quantities of the ectodermal ECM adsorbed to plastic. However, since the (3H)KBH4 method labels only primary amines, it is possible that quantitative differences may exist. In addition, Grinnell and Feld (1981) showed that fibronectin adsorbed to plastic in the presence or absence of other proteins changes the ability of fibronectin to support attachment of BHK cells. Therefore, while these findings are consistent with in situ morphological studies, compositional heterogeneity between cranial ectodermal ECM and trunk ectodermal ECM may cause a particular component within cranial ectodermal ECM or trunk ectodermal ECM to bind to plastic in a configuration that results in altering NC attachment characteristics. Finally, the cranial and trunk ectodermal ECM may not have the ability to reorganize into a substrate completely resembling their in vivo counterpart after the extraction. However, basal- lamina-containing extracellular matrices from different tissues have been extracted, solubilized, and reconsti- tuted and still retain their biological activities (Rojkind et al., 1980; Wicha, 1982; Kleinman et al., 1986).

ECM extracted from the heart was tested for its ability to support NC cell attachment because recent evidence has shown that cranial NC cells enter the embryonic heart and contribute to the formation of the aorticopulmonary septum (Kirby et al., 1983). Many of the ECM molecules of the embryonic heart are similar to those found associated with cranial ectoderm (Markwald et al., 1978; Bolender et al., 1981; Brauer et al., 1985; Mjaatvedt et al., 1987). In the present study, while higher levels were required, the heart ECM transiently supported NC cell attachment. This suggests that particular types of in vivo ECM are more suited for NC cell attachment and may include distant sites eventually populated by NC cells.

280

'".L 30 1

P.R. BRAUER AND R.R. MARKWALD

Cranial N C Attachment

T**

10 260 1000 )cg/ml BSA

T***

h !SO 1000

)cg/ml pFN pqlml PG

pg/ml CrM )cg/ml TrM pg/ml HrM

Fig. 2. Attachment of cranial NC cells to plastic substrates coated with different concentrations of ECM. Substrates were prepared by adsorbing ECM components from solutions containing various concentrations of the test materials (see Materials and Methods). After rinsing, aliquots of 200 cells from a single cranial NC cell suspension were placed on each test substrate and the cells were allowed to attach for 60 minutes at 37°C. At that time, the cultures were rinsed and fixed and the remaining cells were counted. The bar represents the mean cell count from three wells k SEM. The dark horizontal dashed line

represents the mean cell attachment to untreated plastic and the light stippling represents its SEM. Asterisks denote the level of significant difference between the mean cell attachment to untreated plastic and ECM-treated plastic as follows: * * = P < .01 and * * * = P < ,001. The experiment was repeated two times and the results did not vary by more than 15%. BSA, bovine serum albumin; pFN, human plasma fibronectin; PG, bovine nasal cartilage proteoglycans; CrM, cranial ectodermal ECM; TrM, trunk ectodermal ECM; HrM, embryonic heart ECM.

'r 10

0

NEURAL CREST ATTACHMENT TO ECM

Trunk N C Attachment 281

pg/ml BSA pg/ml pFN pg/ml PG

3

1.0 2.5 5.0 1.0 2.5 5.0 pg/ml CrM )cg/ml TrM

Fig. 3. Attachment of trunk NC cells to plastic substrates prepared with different concentrations of ECM after a 60-minute incubation period at 37°C. Substrates were prepared and the attachment assay was conducted as in Figure 2 with the exception that trunk NC cells were substituted for cranial NC cells. Each bar represents the mean cell count from three wells +_ SEM. The dark horizontal dashed line represents the mean cell attachment to untreated plastic and the light

stippling represents its SEM. Asterisks denote the level of significant difference between the mean cell attachment to untreated plastic and ECM-treated plastic as follows: * = P < .05, * * = P < .01. The experiment was repeated three times and the results did not vary by more than 15%. BSA, bovine serum albumin; pFN, human plasma fibronectin; PG, bovine nasal cartilage proteoglycans; CrM, cranial ectodermal ECM; TrM, trunk ectodermal ECM.

282

I 00-

QQ

80-

'"j 80

0

P.R. BRAUER AND R.R. MARKWALD

Cranial N C Attachment at 3 7 O C

I 1 I I I

0 15 30 60

l ime in minutes

Fig. 4. Time course of attachment of cranial NC cells to different ECM-treated plastic substrates at 37°C. Test substrates were prepared by adsorbing ECM components to plastic using concentrations of ECM which gave maximal response in Figure 2. Aliquots of 200 NC cells were placed on each test substrate and allowed to settle for 10 minutes. This was considered the zero time. At various times thereafter, the cultures were rinsed and fixed, and the remaining cells were counted. Each point represents the mean cell count of three wells + SEM.

The results of the NC cell attachment using exogenous ECM components (plasma fibronectin, BSA, cartilage proteoglycan) were similar to those reported by other investigators (i.e., cell attachment to fibronectin was greater than plastic, which was greater than proteoglycan; Newgreen et al., 1982; Erickson and Turley, 1983; Rovasio et al., 1983). Notable, however, was that maxi-

Asterisks denote the level of significant difference between the mean attachment to untreated plastic and ECM-treated plastic as follows: * = P < .05, * * = P < .01, and * * * = P < ,001. The experiment was repeated two times and the results did not vary by more than 15%. pFN, human plasma fibronectin; PG, bovine nasal cartilage proteoglycans; BSA, bovine serum albumin; CrM, cranial ectodermal ECM TrM, trunk ectodermal ECM; HrM, embryonic heart ECM, PL, untreated plastic.

ma1 NC cell attachment occurred on plastic treated with only 1.0 pg/ml of plasma fibronectin, whereas others have reported that much higher concentrations were needed (10-25 pg/mV (Erickson and Turley, 1983; Bou- caut et al., 1984). NC cell attachment to fibronectin decreased as the concentration used to condition the plastic was increased. One explanation may be that

NEURAL CREST A'MACHMENT TO ECM

Cranial N C Attachment at 4" C

50-

0 40- 0 c 0 Q c 2 PD 30- = 0 0 0

20-

I 0-

.... ..... .... r PFN 1 pg/ml

PL

I I i, 15 30 I / I I

60 300

L I I I j-

0 15 30 60 300

Time in minutes

Fig. 5. Time coiirse of attachment of cranial NC cells to different ECM treated plastic substrates at 4°C. The attachment assay was conducted identically to that shown in Figure 4 with the exception that an additional incubation time was included. Each point represents the mean cell count of three wells f SEM. Asterisks denote the level of significant difference between the mean attachment to untreated

concentration-dependent changes in fibronectin conformation occurred during adsorption to plastic (Williams et al., 1982; Erickson and Carrell, 1983). Therefore, it is possible that at lower concentrations fibronectin is adsorbed to plastic in a conformation more suitable for NC cell attachment than when higher concentrations of fibronectin are used to treat the plastic. Another explanation may be that NC cells were exposed to different shearing forces between the assays used by investiga-

283

plastic and ECM-treated plastic as follows: * = P < .05, * * = P < .01. The experiment was repeated two times and the results did not vary by more than 15%. pFN, human plasma fibronectin; PG, bovine nasal cartilage proteoglycans; BSA, bovine serum albumin; CrM, cranial ectodermal ECM; TrM, trunk ectodermal ECM; HrM, embryonic heart ECM; PL, untreated plastic.

tors to remove unattached cells (McClay et al., 1981). One final explanation is that the different results may stem from the fact that most of the previous investiga- tions of NC-cell-substrate interactions have used NC cells grown from a neural tube segment onto the test substrate (primary NC cells) rather than a NC suspension. During the development of a suitable attachment assay, we attempted to use primary NC cells. However, we did not have an adequate means of normalizing the


number of NC cells placed on the test substrates via an outgrowth from a neural tube explant; the extent of neural tube conditioning of each test substrate was un- known; and the number of attached NC cells was too variable for any reliable assessment of NC attachment. Since NC cells migrate as a population in vivo, there is a caveat in using a NC cell suspension and relating these results directly to in vivo NC/ECM interactions. Future experiments comparing the attachment of primary and secondary NC cells to endogenous ECM may determine whether NC cell/ECM interactions are influ- enced by NC cell-NC cell interactions.

Several investigators have shown that human plasma fibronectin is chemotactic for NC cells and that fibronectin-treated plastic or fibronectin-treated collagen enhances avian NC cell attachment in vitro (Greenburg et al., 1981; Sieber-Blum et al., 1981; Newgreen et al., 1982; Erickson and Turley, 1983). In the present study, 1.0 pg/ml of endogenous cranial ectodermal ECM, which contains immunoreactive fibronectin (Brauer and Mark- wald, submitted), supported avian NC attachment as well as 1.0 pg/ml plasma fibronectin even though only a fraction of cranial ectodermal ECM protein was likely to be fibronectin. This suggests that when adsorbed to plastic the endogenous multimolecular complex of cranial ectodermal ECM is more capable of supporting attachment than individual exogenous ECM components.

Since the attachment of cranial and trunk NC cells to cranial ectodermal ECM was greater than to trunk ectodermal ECM, it strongly suggests that these ECM complexes are partly responsible for providing environ- mental cues necessary to direct NC cells into particular pathways at the various axial levels. Recently, it has been shown that ectodermal ECM of axolotl embryos can stimulate the onset of NC cell migration from the neural tube earlier than normal (Lofberg et al., 1985). In addition, the ability of the ectodermal ECM to stimulate NC cell migration was stage dependent and re- gionally restricted (i.e., not freely diffusible effect). Our results suggest that ectodermal ECM is also a factor in directing NC cell migration into particular pathways.

Evidence suggests that subpopulations of NC cells exist at the time NC cells leave the neural tube (Cohen, 1977; Ciment and Weston, 1982; Sieber-Blum and Sie- ber, 1984). Therefore, differences in NC cell attachment to various extacellular matrices could provide a mecha- nism by which subpopulations of NC cells are segre- gated in terms of their migration patterns, proliferation, and phenotypic expression which may have been oper- ating during our attachment assays. In this study, a nonpigmented chicken species was used as the ECM and NC cell source. NC cells which eventually enter the lateral pathway of the trunk are future melanocytes. Therefore, it is possible that only a few of the cells within the NC population from nonpigmented species could utilize trunk ectodermal ECM for attachment whereas NC cells from a pigmented species may have a larger proportion of cells that could attach to the trunk ectodermal ECM. In contrast, the ECM from pigment species and nonpigment species may differ in their com- position or change temporally, allowing more NC cells to enter this pathway. Further experiments utilizing ECM and NC cells from pigmented and nonpigmented species will be necessary to investigate these possibilities.

This study provides evidence suggesting that NC cell migratory pathways may be expressed via the degree to which ectodermal ECM supports NC cell attachment. Work is currently being directed toward isolating the ECM component(s) of cranial ectodermal ECM responsible for enhancing NC cell attachment and to determine whether the effects of these components are negated by the presence of sulfated ECM (e.g., chondroitin sulfate).

ACKNOWLEDGMENTS The authors would like to thank Dr. Edward L. Krug

for the many helpful discussions and Carolyn Snyder for her graphic work. This work was supported in part by grant HL 19136.

LITERATURE CITED Bolender, D.L., W.G. Seliger, R.R. Markwald, and P.R. Brauer (1981)

Structural analysis of extracellular matrix prior to the migration of cephalic neural crest cells. In: Scanning Electron Microscopy/ 1981. O.M. Johari, ed. SEM, Inc., AMF O’Hare, IL, Vol 11, PP. 285- 296.

Boucaut, J.-C., T. Darribere, T.J. Poole, H. Aoyama, K.M. Yamada, and J.P. Thiery (1984) Biologically active synthetic peptides as probes of embrvonic develooment: A comDetitive DeDtide inhibitor of fibronectin ih ib i t s gas tha t ion in amphibian embryos and neural crest cell migration in avian. J . Cell Biol., 99:1822-1830.

Brauer, P.R., D.L. Bolender, and R.R. Markwald (1985) The distribution and spatial organization of the extracellular matrix encoun- tered by mesencephalic neural crest cells. Anat. Rec., 211t57-68.

Bronner-Fraser, M. (1986) Analysis of the early stages of trunk nerual crest cell migration in avian embryos using monoclonal antibody HNK-1. Dev. Biol., 115t44-55.

Ciment, G., and J.A. Weston (1982) Early appearence in neural crest and crest-derived cells of an antigenic determinant present in avian neurons. Dev. Biol., 93:353-367.

Cohen, A.M. (1977) Independent expression of the adrenergic pheno- type by neural crest cells in uitro. Proc. Natl. Acad. Sci. USA, 79:2899-2903.

Crouch, E., G. Balian, K. Holbrook, D. Duskin, and P. Bornstein (1978) Amniotic fluid fibronectin: Characterization and synthesis by cells in culture. J . Cell Biol., 78t701-715.

Erickson, C.A., and E.A. Turley (1983) Substrata formed by combina- tions of extracellular matrix components alter neural crest cell motility in uitro. J. Cell Sci., 61:299-323.

Erickson, H.P., and N.A. Carrel1 (1983) Fibronectin in extended and compact conformations. Electron microscopy and sedimentation analysis. J . Biol. Chem., 258:14539-14544.

Greenburg, J.H., S. Seppa, H. Seppa, and A. T. Hewitt (1981) Role of collagen and fibronectin in neural crest adhesion and migration. Dev. Biol., 87:259-266.

Grinnell, F., and M.K. Feld (1981) Adsorption characteristics of plasma fibronectin in relationship to biological activity. J. Biomed. Mater. Res., 5:363-381.

Haas, R., and L.A. Culp (1982) Properties and fate of plasma fibronectin to the tissue culture substratum. Cell. Physiol., 113:289-297.

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

Hynes, R.O., I.U. Ali, A.T. Destree, V. Mautner, M.E. Perkins, D.R. Singer, D.D. Wagner, and K.K. Smitt (1978) A large glycoprotein lost from the surfaces of transformed cells. Ann. N.Y. Acad. Sci., 312t317-342.

Iwanaga, S., K. Suzuki, and S. Hashimoto (1978) Bovine plasma cold- insoluble globulin: Gross structure and function. Ann. N.Y. Acad. Sci., 312:56-73.

Jilek, F., and H. Hormann (1979) Fibronectin (Cold-Insoluble globulin), VI: Influence of heparin and hyaluronic acid on the binding of native collagen. Hoppe-Seyler’s Physiol. Chemie, 360:597-603.

Johansson, S., and M. Hook (1980) Heparin enhances the rate binding of fibronectin to collagen. Biochem. J., 187:521-524.

Kirby, M.L., T.L. Gale, and D.E. Steward (1983) Neural crest cells contribute to normal aorticopulmonary septation. Science, 220:1059-1061.

Kleinman, H.K., M.L. McGarvey, J.R. Hassell, V.L. Starr, F.B. Can- non, G.M. Laurie, and G.R. Martin. (1986) Basement Membrane complexes with biological activity. Biochemistry, 25:312-318.

Krug, E., R.B. Runyan, and R.R. Markwald (1986) Protein extracts from early embryonic hearts initiate cardiac endothelial cytodiffer-


entiation. Dev. Biol., 112:414-426. Kumarasamy, R., and R.H. Symons (1979) The tritium labeling of

small amounts of protein for analysis by electrophoresis on sodium dodecyl sulfate-polyacrylamide gels. Anal. Biochem., 95:359-363.

Laterra, J., and L.A. Culp (1982) Differences in hyaluronate binding to plasma and cell surface fibronectins. J . Biol. Chem., 257t719-726.

LeDouarin, N.M. (1980) The ontogeny of the neural crest in avian embryo chimeras. Nature, 286.663-669.

LeDouarin, N.M., P. Cochard, M. Vincent, J-P. Duband, G.C. Tucker, M.-A. Teillet, and J.-P. Thiery (1984) Nuclear, cytoplasmic, and membrane markers to follow neural crest cell migration: A com- parative study. In: The Role of Extracellular Matrix in Develop- ment. R.L. Trelstad, ed., 42nd Symposium of the Society for Developmental Biology. Alan R. Liss, Inc., New York, pp. 373-398.

Lofberg, J., A. Nynas-McCoy, C. Olsson, L. Jonsson, and R. Perris (1985) Stimulation of initial neural crest cell migration in the axolotl embryo by tissue graphs and extracellular matrix trans- planted on microcarriers. Dev. Biol., 107:442-459.

Markwald, R.R., T.P. Fitzharris, H. Bank, and D.H. Bernanke (1978) Structural analyses on the matrical organization of glycosaminoglycans in developing endocardial cushions. Dev. Biol., 62:292- 316.

McClay, D.R., G.M. Wessel, and R.B. Marchase (1981) Intercellular recognition: Quantitation of initial binding events. Rw. Natl. Acad. Sci. USA, 78:4975-4979.

Mjaatvedt, C.M., R. Lepera, and R.R. Markwald (1987) Myocardial specificity for initiating endothelial-mesenchymal cell transition in embryonic chick heart correlates with a particulate distribution of fibronectin. Dev. Biol., 119:59-67.

Newgreen, D.F., and J.P. Thiery (1980) Fibronectin in early avian embryos: Synthesis and distribution along migratory pathways of neural crest cells. Cell Tissue Res., 211:269-291.

Newgreen, D.F., I.L. Gibbins, J . Sauter, B. Wallenfels, and R. Wurtz (1982) Ultrastructural and tissue-culture studies on the role of fibronectin, collagen, and glycosaminoglycans in the migration of neural crest cells in the fowl embryo. Cell Tissue Res., 221:521- 549.

Noden, D.M. (1978a) The control of avian cephalic neural crest cell cytodifferentiation. I. Skeletal and connective tissues. Dev. Biol., 67:296-312.

Noden, D.M. (1978b) The control of avian cephalic neural crest cell cytodifferentiation. 11. Neural tissues. Dev. Biol., 67t313-329.

Noden, D.M. (1984) Craniofacial Development: New views on old prob- lems. Anat. Rec., 208:l-13.

Peterson, G.L. (1977) A simplification of the protein assay method of Lowry et al. which is more generally applicable. Anal. Biochem., 83:346-356.

Pratt, R.M., M.A. Larson, and M.C. Johnston (1975) Migration of cranial neural crest cells in a cell-free hyaluronate-rich matrix. Dev. Biol., 44:298-305.

Rojkind, M., Z. Gatmaitan, S. Mackensen, M.-A. Giambrone, P. Ponce, and L.M. Reid (1980) Connective tissue biomatrix: Its isolation and utilization for long-term cultures of normal rat hepatocytes. J. Cell Biol., 87:255-263.

Rovasio, R.A., A. DeLouvee, K.M. Yamada, R. Timpl, and J.P. Thiery (1983) Neural crest cell migration: Requirements for exogenous fibronectin and high cell density. J. Cell Biol., 96:462-473.

Sieber-Blum, M., and F. Sieber (1984) Heterogeneity among early neural crest cells. Dev. Brain Res., 14:241-246.

Sieber-Blum, M., F. Sieber, and K.M. Yamada (1981) Cellular fibronectin promotes adrenergic differentiation of quail neural crest cells in uitro. Exp. Cell Res., 133:285-295.

Teillet, M.A., and N. LeDouarin (1970) La migration des cellules pig- mentaires etudie par la methode heterospecifiques de tube nerveus chez I’embryon d’Oiseau. C.R. Acad. Sci. Ser. D., 270:3095-3098.

Thiery, J.P., J.P. Duband, and A. DeLouvee (1982) Pathways and mech- anisms of avian trunk neural crest cell migration and localization. Dev. Biol., 93:324-343.

Weston, J.A., M.A. Derby, and J.E. Pintar (1978) Changes in the extracellular environment of neural crest cells during early migration. Zoon, 6:103-113.

Wicha, M.S. (1982) Growth and differentiation of rat mammary epithe- lium on mammary gland extracellular matrix. In: Extracellular Matrix. S. Hawkes, and J.L. Wang, eds. Academic Press, New York, pp. 309-314.

Williams, E.E., P.A. Janney, P.D. Ferry, and D.F. Mosher (1982) Con- formational states of fibronectin. Effects of pH, ionic strength, and collagen binding. J. Biol. Chem., 257:14973-14978.

Yamada, K.M., and D.W. Kennedy (1979) Fibroblast cellular and plasma fibronectins are similar but not identical. J. Cell Biol., 80:492-498.

Yamada, K.M., and D.W. Kennedy (1984) Dualistic nature of adhesive protein function: Fibronectin and its biological active peptide frag- ments can autoinhibit fibronectin function. J . Cell Biol., 99:29-36.

Documents

Attachment of neural crest cells to endogenous extracellular matrices