Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
DEVELOPMENTAL BIOLOGY 107,66-74 (1985) Three Cell Recognition Changes Accompany the Ingression of Sea Urchin Primary Mesenchyme Cells RACHEL Department D. FINK AND DAVID R. MCCLAY’ of Zoology, Duke University, Received February Durham, North 23, 1984 accepted in revised fwm Carolina 2~06 July 27, 1984 At gastrulation the primary mesenchyme cells of sea urchin embryos lose contact with the extracellular hyaline layer and with neighboring blastomeres as they pass through the basal lamina and enter the blastocoel. This delamination process was examined using a cell-binding assay to follow changes in affinities between mesenchyme cells and their three substrates: hyalin, early gastrula cells, and basal lamina. Sixteen-cell-stage micromeres (the precursors of primary mesenchyme cells), and mesenchyme cells obtained from mesenchyme-blastula-stage embryos were used in conjunction with micromeres raised in culture to intermediate ages. The micromeres exhibited an affinity for hyalin, but the affinity was lost at the time of mesenchyme ingression in vivo. Similarly, micromeres had an affinity for monolayers of gastrula cells but the older mesenchyme cells lost much of their cell-to-cell affinity. Presumptive ectoderm and endoderm cells tested against the gastrula monolayers showed no decrease in binding over the same time interval. When micromeres and primary mesenchyme cells were tested against basal lamina preparations, there was an increase in affinity that was associated with developmental time. Presumptive ectoderm and endoderm cells showed no change in affinity over the same interval. Binding measurements using isolated basal laminar components identified fibronectin as one molecule for which the wandering primary mesenchyme cells acquired a specific affinity. The data indicate that as the presumptive mesenchyme cells leave the vegetal plate of the embryo they lose affinities for hyalin and for neighboring cells, and gain an affinity for fibronectin associated with the basal lamina and extracellular matrix that lines the blastocoel. @J 19x5 Academic PRSS. hc. The environment of the primary mesenchyme cells changes as they move from the wall of the blastula into the blastocoel. They lose contact with the extracellular hyaline layer and with neighboring cells. The cells ingress through the basal lamina, and begin a series of migrations on the internal face of that structure. Okazaki (1960) observed the movements of the primary mesenchyme cells after ingression and showed that they follow a distinct pattern prior to and during spicule formation. It was later demonstrated that isolated micromeres follow the same behavioral patterns in culture as their in vivo counterparts, including production of spicules (Okazaki, 1975). We have studied the cell surface properties of micromeres and their descendents using cells from two sources. Cells cultured in vitro provide a highly enriched source of mesodermal cells that can be harvested at any precise time during development, though it is necessary to show that cultured cells behave as their counterparts in viva. Micromeres and primary mesenthyme cells were also isolated and studied directly from embryos. The use of these uncultured cells allowed us to measure recognition properties at the two ends of the behavioral spectrum, while the use of cells in vitro permitted us to examine recognition changes at the precise time of primary mesenchyme cell ingression. INTRODUCTION The ingression of primary mesenchyme cells is the earliest recognizable morphogenetic event of gastrulation in the sea urchin embryo. Several detailed studies have documented the sequence of events in primary mesenchyme migration. For example, Gustafson and Kinnander’s time-lapse microcinematographic study of sea urchin development focused on the extensive migrations of the primary mesenchyme cells (1956), and demonstrated the sudden change in active motility of these cells at ingression. Examinations of ingressing primary mesenchyme cells by electron microscopy revealed a dramatic alignment of microtubules as the cells changed shape (Gibbons et aL, 1969). More recently, Katow and Solursh (1979) have studied the ultrastructure of these movements, and described the fibrous nature of blastocoelic material that may serve as a substrate for cell movement. These descriptive studies form the necessary framework for functional studies on the basis of morphogenetic movements. This paper examines several cell recognition properties to determine their contributions to mesenchyme cell ingression and migration. ’ To whom all correspondence 0012-1606/G Copyright All rights should be addressed. $3.00 0 1985 by Academic Press, Inc. of reproduction in any form reserved. 66 FINK AND MCCLAY Sea Urchin Our working hypothesis is based on a model proposed by Gustafson and Wolpert (1967) which suggested that changes in cell-cell and cell-matrix adhesions are components of the morphogenetic movements of primary mesenchyme cells. To test this hypothesis we have used a sensitive cell binding assay to measure the affinity of sea urchin primary mesenchyme cells for other embryonic components. We have previously reported a developmentally regulated loss in affinity for hyalin by postingressive mesenchyme cells at a time corresponding to their normal release from the hyaline layer (McClay and Fink, 1982). This report shows that there are additional developmentally regulated changes in the adhesive behavior of mesenchyme cells with nonmesodermal cells and with isolated basal laminar components. MATERIALS AND METHODS Animals and Gametes Arbaha punctulata were obtained from the Duke University Marine Laboratory and Strongylocentrotus purpuratus were purchased from West Coast Suppliers. Gametes were shed by injection with 0.5 M KCl. Eggs were fertilized with dilute sperm suspensions, and fertilization membranes removed with P-aminobenzoic acid (Paba, 10 mM) (McClay and Fink, 1982). Embryos were grown in seawater filtered through a 0.45-pm pore-size filter (Millipore) (MSW) containing penicillin (l-2 pg/ml) and streptomycin sulfate (l-2 pg/ml) (Sigma), or in calcium-free seawater (CF) through the 16-cell stage, at 15°C (Strongylocentrotus) or at 2224°C (Arbacia). Cell Culture and Dissociation Mesenchymal cells were isolated by two methods. In order to obtain mesenchyme cells prior to complete ingression, 16-cell-stage micromeres were isolated after the method of Okazaki (1975). Briefly, fertilized eggs were placed into CF after sperm addition and fertilization membrane removal. The embryos were grown to the 16-cell stage in CF, then they were gently dissociated by swirling and pipetting. The dissociated embryos were layered on a discontinuous sucrose gradient (4%,8% in CF), and the micromeres collected from the 4-8% interface. Micromeres were grown in MSW supplemented with 3% horse serum and antibiotics. Mesenchymal differentiation was monitored microscopically, and ingressive stage and wandering stage cultured mesenchyme cells were harvested at the appropriate times by gentle pipetting. Alternatively, mesenchyme cells were isolated from mesenchyme blastula stage embryos after the methods of McClay and Marchase (1979) and Harkey and Whiteley (1981). This isolation Cell Recognition Changes 67 involves dissociating mesenchyme blastulae in a CF: hyaline extraction medium (HEM) mixture, which gently removes the noningressed blastomeres. The resulting basal laminar bags contain isolated primary mesenchyme cells. These are gently sheared and the cells collected by centrifugation. Embryos at the blastula stage and beyond were dissociated by treatment for l-5 min with HEM: 0.3 M glycine, 0.3 M NaCl, 10 mM KCl, 10 mM MgS04, 10 mM Tris, 2 mM EGTA (ethylene glycol bis@aminoethyl ether)-NJ’-tetraacetic acid), pH 8.0 (McClay and Fink, 1982). The hyaline layer was solubilized by HEM and the embryos were easily dissociated by gentle pipetting. Single-cell suspensions were separated from undissociated clumps by pouring through 28-pm Nitex mesh (Tetko). Cell suspensions were washed with MSW and were found to exclude trypan blue (greater than 90% of a suspension) and, as a second test of cell viability, the cells aggregated well in rotary culture. Ectoderm cells were separated from endoderm cells by procedures described previously (McClay and Marchase, 1979). Basal Lamina Isolation Basal lamina components were isolated as intact basal laminar bags from embryos at the midblastula stage and beyond. Embryos were lysed in 10 mM NaHC03, 0.01% Triton X-100 (Wessel et aL, 1984). Bags were collected and washed by centrifugation, and then treated with DNase (30 pg/ml) (Sigma) to remove trapped nuclei. Alternatively, embryos were dissociated with HEM and the basal lamina bags collected by gentle centrifugation (McClay and Marchase, 1979; Harkey and Whiteley, 1981). In both cases the isolated bags were sheared into fragments. Several experiments used the following basal laminar components: fibronectin (Sigma), laminin (gift of H. Kleinman), collagen (mixture of types I-IV, Sigma), chondroitin sulfate (Sigma), and hyaluronic acid (Sigma). Cell-Binding Assay The cell-binding assay used has been described previously (McClay et al. 1981; McClay and Fink, 1982). Briefly, proteins or cell monolayers are adsorbed onto the bottom of polyvinyl chloride flat-well microtiter plates. Aliquots of radiolabeled probe cells are then added to each well and the plates sealed. The probe cells are gently centrifuged into contact with the substrate. The microtiter plates are then inverted, and can be centrifuged in this inverted position to add a dislodgement force of greater than lg. The microtiter plates are then quick-frozen at -7O”C, and the bottoms of the wells (containing the substrate layer and any 68 DEVELOPMENTALBIOLOGY radiolabeled probe cells that remained bound during the treatment) are clipped off for liquid scintillation counting. Three different substrates were used in the binding assay. (1) Hyalin, the major protein component of the hyaline layer, was purified from early cleavage stage embryos (McClay and Fink, 1982) and directly adsorbed onto the microtiter wells (lo-20 pg/well). (2) Fragments of complete basal laminae were centrifuged onto the microtiter wells (approximately 30 pg protein/well). Purified components of basal lamina were adsorbed directly onto the microtiter wells (lo-20 pg/well). Protein content of the adsorbed substrates was measured by a miniaturized Bradford assay. (3) Cell monolayers were generated by treating the microtiter plates with a 21 solution of poly+lysine (2 mg/ml): glutaraldehyde (3% in MSW), followed by extensive washings with MSW. An aliquot of a single-cell suspension was then added to each well (9 X lo5 cells/well), and the microtiter plates were centrifuged at 200~ to force the cells onto the bottom of the well. Confluency of monolayers was determined by microscopic observation. A series of controls on the bioassay showed that the cells remained intact (there was no radioactive material released into the supernatant, and the ratio between lz51-surface labeled: 3H-internally labeled cells did not change after cells were removed from the substrate). Monolayers were not released from the wells as determined by initial controls in which the monolayer cells were radiolabeled. The substrate itself could be blanked by washing a well with 10% fetal calf serum in MSW to assure that cells bound to monolayers or matrix materials rather than any free surface that might exist on the well bottom (8% background binding resulted if cells were exposed to an entire well surface treated with 10% fetal calf serum; these same cells bound at a level of 95100% to wells treated with polyL-lysine alone). Micromeres, cultured mesenchyme cells, and newly isolated mesenchyme cells were radiolabeled for use as probe cells in the binding assays. To radiolabel micromeres, fertilized eggs were placed in [3H]leucine (1 &i/ml) in CF until the 16-cell stage (approx 3 hr for Arbacia, 5 hr for Strongylocentrotus). [3H&eucine (1 &i/ml) was added to cultured mesenchyme cells or to morula stage embryos 4 hr before use as probe cells. Labeled cells were washed extensively in MSW, and added to the microtiter wells on ice in a total volume of 0.3 ml. The same number of cells (approx 105; O.Ol0.05 cpm/cell) were then added to each well and the plates sealed. Probe cells were centrifuged onto the substrates at lo-20g. After 15 min incubation on ice, the plates were inverted and centrifuged while in the inverted position to add a dislodging force greater VOLUME107,1985 than lg to the cells (McClay et al., 1981). Controls each day included blanked wells for which the cells had no affinity (3-8% background binding). To estimate the strength of adhesion, test cells were centrifuged away from a substrate at varying centrifugal forces. A plot of percentage binding versus relative centrifugal force (RCF) applied was obtained for each cell type on each of the different substrates. These graphs were used to estimate the centrifugal force necessary to remove 50% of the test cells from the substrate at the different developmental stages (McClay et aL, 1981). This centrifugal dislodgement force (Fn) was calculated from the equation FD = (pm11 - Pmedium) X Ke11 X RCF in which Peel1is the specific density of the cell (determined for echinoderm cells to be 1.256 g/cm3); Pmedium is the specific density of the medium (MSW, 1.024 g/cm3); and V,,, is the cell volume. A diameter of 16 pm was used for l&cell-stage micromeres, and 10 pm for wandering mesenchyme cells. This calculation considers size difference between cells of different ages. In some cases, the maximum binding achieved at a removal force of lg was lower than 50%. For these interactions, FD was estimated for the lg binding value. Mono&ma1 Immunojluorescence Several primary mesenchyme cell-specific monoclonal antibodies have been generated (McClay et aL, 1983) and were used in this study. The monoclonal antibodies were raised against Lytechinus variegates gastrula membranes (McClay et aL, 1983; Wessel et aL, 1984). These antibodies were used in indirect immunofluorescent assays on fixed embryonic tissues. To determine the staining pattern in whole embryos, embryos were fixed in Bouin’s fixative, dehydrated in an ethanol series, embedded in paraffin, and serially sectioned at 5 pm (Wessel et ah, 1984). The mounted sections were incubated with monoclonal supernatant in a humid chamber. The sections were rinsed in phosphate-buffered saline containing 0.05% Tween-20 (PBS-Tw20), and incubated with fluorescein isothiocyanate (FIT(?)conjugated rabbit anti-mouse Ig. The sections were then rinsed with PBS-Tw20, converslips were added, and they were viewed under a Leitz fluorescent microscope equipped with epiluminescence. For cultured mesenchyme cells, cells were raised on glass coverslips. At the appropriate time the cells were fixed in cold acetone for 10 min, and air dried. Cells were then stained with the same series of monoclonal antibody followed by FITC-conjugated secondary antibody as described above, and viewed under fluorescent light. FINK AND MCCLAY Sea Urchin Cell Recognition Changes Controls included sections without the monoclonal but with the secondary antibody, substitution of myeloma supernatant for antibody, and omitting the FITC secondary antibody to screen for autofluorescence. Each of these was negative. A positive control using a polyclonal anti-sea urchin serum and other monoclonals with different specificities (McClay et a& 1983) was included in most experiments to assure specificity of the mesenchyme antibody. RESULTS Development of Micromeres in Culture Cultured cells were used in the cell binding experiments to identify and determine the timing of changes in cell recognition properties. In order to use such cells it was necessary to establish the degree to which they mimicked growth and differentiation in vivo. Sixteencell-stage micromeres were isolated and raised in culture, and exhibited the same morphological and behavioral patterns seen and previously described for cells in vivo (Okazaki, 1975). Newly isolated micromeres underwent a number of divisions, forming small clusters of cells. At the precise time of mesenchyme 69 ingression in control embryos, cultured mesenchyme cells began pulsatory movements, extended transitory pseudopodial cell processes, and became motile on the culture dish. At the time corresponding to midgastrula stage, cultured mesenchyme cells sent out filopodia which overlapped and fused to form syncytial spicule envelopes, within which mineralization occurred. Small CaC03 crystal rudiments appeared, and the spicule grew in vitro as the mesenchyme cells moved away from the initial crystal, causing elongation. Thus, cells cultured in vitro exhibited the same behavioral changes as the primary mesenchyme cells in vivo. To further assess the development of primary mesenchyme cells in vitro we examined the expression of cell surface markers. Monoclonal antibodies against primary mesenchyme cells were screened (McClay et al, 1983; Wessel et al, 1984). In sectioned whole embryos, no fluorescent staining was seen until the mesenchyme blastula stage when primary mesenchyme cells expressed the antigen (Fig. 1). As development continued, the staining remained specific for primary mesenchyme cells and for the envelope covering the spicules they secreted. When these antibodies were used to monitor cell FIG. 1. Monoclonal antibody staining in viwo and in vitro. Indirect immunofluorescence using a monoclonal antibody specific for primary mesenchyme cells shows that at the precise time of antigen appearance in viva, the cultured cells begin to show strong fluorescent staining. At mesenchyme blastula stage, the antibody recognized mesenchyme cells that had completed ingression (a, b). This fluorescence remains specific for primary mesenchyme cells and their products during development. At pluteus stage, the antigen binds to primary mesenchyme cells and to the larval spicules (c, d). In vitro cells first show fluorescent staining when they become migratory (e, f). This corresponds to the mesenchyme blastula stage in viva. In the culture system, the antibody continues to recognize the mesenchyme cells and the envelope surrounding the spicule (g, h). Scale bars: (a) 15 pm; (c) 20 pm; (e) 8 pm; (g) 12 pm. 70 DEVELOPMENTALBIOLOGY surface antigens on cultured mesenchyme cells, the cells became fluorescent as they became migratory. At that time, which corresponds to ingression in viva, cells began to react strongly with the antibodies (Fig. 1). Throughout spicule development in vitro, the staining pattern was identical to that in whole embryos. The coincidental appearance of these mesodermalspecific antigens on cultured mesenchyme cells suggests that their cell surface is comparable to this cell type in viva. Thus because the timing of behavioral events, the antigen expression, and the morphological changes were similar, we felt we could use the cultured cells in the cell-binding assay with confidence of cell stage. Measurement of Cell-Cell and Cell-Substrate A$inity To quantify the adhesive interactions exhibited by the primary mesenchyme cells, binding assays were performed that measured affinities between cells or between cells and substrate materials (McClay et ak, 1981). The properties of the cell-binding assay are shown in Fig. 2. Labeled midgastrula probe cells were tested against gastrula monolayers using a series of centrifugal forces to separate probes from attached monolayers. At 4°C the percentage binding versus relative centrifugal force profile shows a decline in cell-to-cell binding over the range of centrifugal forces applied. At low centrifugal forces, 100% of the test cells remain bound to the monolayers. A removal force I I 25 50 100 RCF 200 500 1000 FIG. 2. Percentage binding versus relative centrifugal force (RCF). Binding of dissociated gastrula cells to gastrula monolayers at two different temperatures. Percentage binding over a range of relative centrifugal removal forces was determined by the cell binding assay (see Materials and Methods). The 24°C binding profile represents metabolically stabilized cell-cell contacts (see text). Data reported in this paper were obtained from assays performed at 4°C. The 4’C RCF value giving 50% binding (2OOg) can be used to estimate a median dislodgement force (McClay et aL, 1981). VOLUME 107, 1985 of 2009 reduces the number of test cells bound to about 50%. By 10009 all but 20% of the test cells have been removed. In contrast, when the probe cells and monolayers were incubated at 24°C for 15 min, the probe cells were much harder to remove. One-hundred percent of the probe cells remained bound to the monolayers even at removal forces of 200g. Binding was only reduced to 65% at the maximum removal force of 10009. Thus even a short incubation at 24°C appears to stabilize the adhesive interactions between probe and monolayer cells. Cell-To-Cell Afinity Changes To examine the relationship of developing mesenthyme cells with other cells of the embryo, confluent monolayers of dissociated early gastrula cells were used as substrates in a series of assays. Radiolabeled mesenchyme probe cells were added to monolayercovered microtiter wells at 4°C and centrifuged into contact with the substrate at log. The plates were inverted and centrifuged in this inverted position at a relative centrifugal force of 50g. Table 1 shows that prior to mesenchyme cell ingression, newly isolated or cultured micromeres had an affinity for gastrula cells. This affinity was reduced in migratory-stage mesenthyme cells. The decrease in cell-to-cell affinity was exhibited by postingressive mesenchyme cells of both Arbacia and Str~ylocentrotus. By contrast, when other cells of the embryo were tested against the same gastrula monolayers, a high degree of binding was measured. All through gastrulation this affinity was retained both by isolated ectoderm and by endodermal cells. Thus only descendents of micromeres lost an affinity for other cells and this loss occurred at the mesenchyme blastula stage. From profiles of binding versus centrifugal force applied, Table 3 shows that 16-cell-stage micromeres resisted a removal force of about 7 X 10m5dyn per cell. Migratory-stage mesenchyme cells lost the ability to bind and the force dropped to less than 10m7dyn per cell (the median number of cells bound at lg was less than 50%). The calculated binding affinity for gastrula cells (ectoderm and endoderm) remained unchanged over this same developmental interval. The Interaction between Cells and the Basal Lamina The basal lamina appears during cleavage as the blastocoel forms (Wessel et al., 1984). Each blastomere is in contact with the basal lamina which thickens and gains structural integrity by the mesenchyme blastula stage (Gibbons et aZ., 1969). Early mesodermal cells are in association with the outer face of the basal lamina. Upon ingression into the blastocoel, primary mesenchyme cells establish close contact with the FINK AND MCCLAY Sea Urchin TABLE 1 AFFINITY TO GASTRULA MONOLAYERS” Percentage Probe cell type 16-Cell-state micromeres Migratory mesenchymed Uncultured mesenchymed Dissociated blastulae Dissociated gastrulae Arbucia 82 22 27 66 86 i 11 (5)C f 9 (4) f 11 (2) (1) + 7 (3) bindingb S. purpuratus 88 + 9 (3) 46 31 3 (2) ND” ND ND a Monolayers made from single-cell suspensions generated upon dissociation of early gastrulae (predominantly ectoderm and endoderm cells, with few mesodermal cells). bBinding measured in microtiter well assay (see Materials and Methods). A removal force of 50g was applied; assays were performed at 4°C. Arbacia cells were tested against Arbacia monolayers; S. pzqwuratus cells against S. purpuratus monolayers. ‘Data presented are means + SD of binding assays performed on different days. Numbers in parentheses are the number of experiments. Values for each experiment are the mean binding for at least six replicates. Within an experiment, errors were 10% or less. d Cells were isolated as 16-cell-stage micromeres and raised in vitro, or isolated from mesenchyme blastulae (uncultured). e Not determined. inner face of the basal lamina and the extracellular matrix which serve as a substrate for their subsequent migrations (Okazaki, 1965). To determine whether mesenchyme cells undergo affinity changes for the basal lamina, preparations of basal lamina, isolated as bags from late mesenchyme blastula stage embryos and then fragmented, were bound as substrates to microtiter wells of assay plates. Sixteen-cell-stage micromeres of both Arbacia and Strongylocentrotus exhibited some degree of binding to the basal laminar fragments (Table 2). Arbacia micromeres that had been cultured to the early blastula stage showed no significant increase in binding affinity for basal laminae. At the migratory stage, the strength of the measurable binding affinity increased for the mesenchyme cells that normally use the basal lamina as a migratory substrate. Dissociated blastula cells and gastrula cells (ectoderm and endoderm) both showed an affinity for basal lamina that did not change with developmental time. The binding strength of early micromeres was calculated to be on the order of lo-? dyn per cell, and for wandering mesenchyme cells it was about 10e5dyn per cell (Table 3), representing a large increase in binding affinity of wandering mesenchyme cells for the basal lamina. The increase appears to occur very rapidly in development; the duration of primary mesenchyme cell ingression in vivo is 1 hr, and cells cultured to just prior to ingression do not demonstrate this increased affinity. Cells cultured to just after the ingression, or primary mesenchyme cells isolated after ingression in Cell Reco,qnitim 71 Changes vivo both demonstrate the increased affinity for basal lamina. Characterization of sea urchin basal lamina has identified several components that are cross reactive with those of vertebrate basal laminae (Spiegel et al, 1980; Wessel et al, 1934). To determine whether the measured increase in binding affinity for complete basal lamina was due to a subset of these components, a series of binding assays was performed. The binding affinity of l&cell-stage micromeres and migratory mesenchyme cells were similar for laminin, collagen, chondroitin sulfate, and hyaluronic acid (Table 4). When micromeres and wandering mesenchyme were tested against fibronectin, however, the older cells had a higher affinity for this component. This increased affinity for fibronectin was exhibited by micromeres cultured through the wandering mesenchyme cell stage as well as mesenchyme cells newly isolated from postmesenchyme blastula stage embryos (Table 4). The horse serum used for culture had no apparent effect on the cultured cells because the affinity change correlated with time of mesenchyme blastula stage rather than inclusion of, or length of time in the presence of, serum. Cell-Hgalin Binding Strength In a previous study we examined the relationship of mesenchyme cells to hyalin (McClay and Fink, 1982). Sixteen-cell-stage micromeres and presumptive primary mesenchyme cells showed a high degree of binding to TABLE 2 AFFINITY TO BASAL LAMINA~ Percentage Arbacia Probe cell type 16-Cell-stage micromeres Morula stage mesenchyme Migratory mesenchymed Dissociated blastulae Dissociated gastrulae cellsd 19 23 55 23 20 + 12 (4)c (1) + 6 (3) + 4 (2) -t 8 (2) bindingb S. purpuratus 20 + 9 ND” 58 + 10 47 52 (2) (2) (1) (1) a Substrates were complete, isolated basal laminae. bBinding measured in microtiter well assay (see Materials and Methods). A removal force of 50g was applied; assays were performed at 4°C. A&a&a cells were tested against Adacia basal lamina; S. pqmratus cells against S. purpratus basal lamina. ‘Data presented are means f SD of binding assays performed on different days. Numbers in parentheses are the number of experiments. Values for each experiment are the mean binding for at least four, and usually six, replicates. Within an experiment, errors were 10% or less. ‘Cells were isolated as 16-cell-stage micromeres and raised in culture to the morula stage (prior to ingression) or to the migratory stage (after ingression was completed). e Not determined. 72 DEVELOPMENTALBIOLOGY TABLE 3 ESTIMATESOF BINDING VOLUME107, 1985 STRENGTH” Force* required to dislodge cells from: Cell type Hyalin Monolayers Basal lamina 16-Cell-stage micromeres Migratory-stage mesenchyme cells Gastrula ectoderm and endoderm 5.8 x 10-5 1.2 x 1o-7 (15%) 5.0 x 1o-6 6.8 x 1o-5 1.2 X 1O-7(46%) 5.0 x w5 4.8 x 1o-7 (20%) 1.5 x 1o-5 5.0 x 10-7 “As measured by a series of adhesion assays performed under different centrifugal removal forces. See Materials and Methods, and Fig. 2. *Dislodgement force, in dynes, is calculated for that centrifugal force that removes 50% of the test cells from the substrate. ‘Number in parentheses is the maximum binding achieved (when less than 50%), and the binding strength estimate, in dynes, is for that value. hyalin that was lost at the time of mesenchyme ingression. Table 3 shows that 16-cell-stage micromeres were dislodged from hyalin at about 10e5 dyn per cell; wandering mesenchyme cells were released at about lo-’ dyn per cell. DISCUSSION The results of the binding studies show that primary mesenchyme cells, at the mesenchyme blastula stage, TABLE 4 AFFINITY TO ISOLATED Substrate” Fibronectin Laminin Collagen Hyaluronic acid Chondroitin sulfate BASAL LAMINA COMPONENTS Probe cells* Micromeres’ Cultured mesenchyme Micromeres Uncultured mesenchyme Micromeres Cultured mesenchyme Micromeres Cultured mesenchyme Micromeres Cultured mesenchyme Micromeres Cultured mesenchyme % Binding” 20 t 72 * 18 + 55 2 26 k 31 + 28 It 31 + 41 + 34 46 f 55 9 (4)d 17 (3) 7 (2) 13 (3) 10 (3) 21 (2) 14 (3) 15 (2) 9 (2) (1) 8 (2) (1) ‘Sources and concentrations of substrates are described under Materials and Methods. bThree different probe cells were used: (1) 16-cell stage micromeres (Micromeres), (2) mesenchyme cells isolated from mesenchyme blastula stage embryos (Uncultured mesenchyme), and (3) mesenchyme cells that were isolated as micromeres and raised in culture (Cultured mesenchyme). “Binding measured in microtiter well assay (see Materials and Methods section). A removal force of 5Og was applied; assays were performed at 4°C. ‘Data presented are means k SD of binding assays performed on different days. Numbers in parentheses are the number of experiments. Values for each experiment are the mean binding for six replicates. Within an experiment, errors were 10% or less. ePresence or absence of exogenous serum did not affect the results. lose an affinity for hyalin and for other cells. At the same time they gain a measurable affinity for basal lamina that may be specific for fibronectin (Fig. 3). Ectodermal and endodermal cells, under the same conditions, showed no such developmentally regulated changes in affinity (i.e., dissociated blastulae as early nonmesenchymal test cells, and combined ectoderm and endoderm from the gastrula stage, displayed no changes in their affinities for hyalin, embryonic monolayers, or basal lamina). Thus the changes in initial binding behavior appear to reflect specific affinity changes of primary mesenchyme cells at the time of ingression. To measure the timing of the affinity changes two cell populations were used. Micromeres were isolated from the 16-cell stage and primary mesenchyme cells were isolated from the mesenchyme blastula stage. To examine the precision of the affinity changes we used cultured micromeres. Several lines of evidence suggested that the cultured mesenchyme cells are comparable to those in whole embryos, and that the cultured cells undergo the affinity changes just as ingression occurs in vivo. First, the behavior and FIG. 3. Changes in adhesive interactions with developmental time. As presumptive primary mesenchyme cells (P) begin to ingress to the interior of the embryo, they lose affinities for hyalin (H) and for neighboring blastomeres, and gain an affinity for the basal lamina (BL). Nonmesenchymal blastomeres (E) exhibit no such developmentally regulated changes in affinity. FINK AND MCCLAY Sea Urchin timing of cultured mesenchyme cells paralleled normal development. Second, monoclonal antibodies specific for postingressive primary mesenchyme cells recognized cultured cells at the precise stage this antigen was expressed in vivo. Third, cells grown in vitro to the primary mesenchyme cell stage had the same altered affinity properties as the corresponding cells obtained from whole embryos (Table 4; McClay and Fink, 1982). Fourth, micromeres, in the presence or absence of serum, or cells raised in culture (2 to 6 hr) in the presence of serum for any time up to the ingression event behaved in the same way as uncultured micromeres. This argues that the presence of serum in the culture medium has no artifactual effect on the affinity changes. Thus it would appear that cells grown in culture accurately reflect the developmental changes of their counterparts in vivo. We therefore conclude that the affinity changes of primary mesenchyme cells occur coincident with mesenchyme cell ingression. The cell-binding assays in this study last about 10 min each and were carried out at 4°C in an effort to restrict observations to initial binding events only. Under these conditions other adhesion-associated events (such as cell junction formation) probably do not occur. Quantitation of the dislodgement forces showed that cells had an affinity for one another in the range of 10-5-10-’ dyn per cell or less. When cells were allowed to warm to 24°C for even a brief time, a dramatic increase in binding affinity was measured (Fig. 2). This high level of binding probably represents any of a number of stabilization events such as the formation of cell junctions, cell shape change to increase surface area of contact between cells, or secretion and adhesion to extracellular matrix materials. Any of these events might also participate in primary mesenchyme morphogenesis but in the present context even the earliest events in the adhesive process reflect potentially important morphogenetic properties. In the binding studies cell orientation cannot be controlled when the cells settle into contact with the substrate. Thus, if blastomeres have only a portion of their surface area that can bind to the substrate, such blastomeres as a population would have a reduced affinity for the substrate relative to blastomeres that had binding components covering the entire surface. It is possible, therefore, that some of the affinity changes observed could be due to some kind of a mosaic restriction of surface binding sites. The alternative to this possibility is a change in the identity or quantity of molecular components on the surface of the primary mesenchyme cells at ingression. In either event, the mesenchyme cells increase in their affinity for one substrate while decreasing in their affinity for two Cell Recognition Changes 73 other substrates. Thus, at least two changes must occur and these changes must be at the surface of the primary mesenchyme cells since the experimental substrates used (hyalin, cell monolayers, and basal lamina) were the same throughout. The developmentally regulated changes in cell-cell and cell-matrix binding reported here support the Gustafson and Wolpert (1967) model of morphogenesis. The sudden change in cell affinities observed here may not be unique to sea urchin primary mesenchyme cells. Other organisms, for example, display similar epithelial-mesenchymal morphogenetic shifts. In vertebrate embryos neural crest cells break away from neighboring cells in the neural folds as they attain the migratory morphology of mesenchyme cells (Tosney, 1978). Also, at gastrulation in the chick, mesenchyme cells break away from the epiblast of the primitive streak and undergo an epithelial-mesenchymal change (Solursh and Revel, 1978). For these cell types, and for others, changes in cell affinities may occur at specific times to represent a general mechanism that precedes or accompanies morphogenetic movements. Mechanistically, movement of a cell on a substratum can be modeled as the transient formation and breakage of localized bonds. The initial affinities measured by the binding assay may represent such bonds. If this were the case then the observed specific increase in binding to fibronectin may have importance for the mechanism by which primary mesenchyme cells move on the wall of the blastocoel. Several studies have correlated fibronectin location with coincidental morphogenetic movement. In amphibian embryos, for example, fibronectin first appears on the roof of the blastocoel along the pathway of mesoderm cell migration (Boucaut and Darribere, 1983; Lee et cd., 1984). In the chick fibronectin has been detected along the path of mesoderm migration in the primitive streak (Critchley et ah, 1979). Later in development fibronectin has been observed in the pathway of neural crest migration (Rovasio et al., 1983). In the sea urchin fibronectin has been localized, by immunofluorescence, to the surface of the blastocoel and to primary mesenchyme cells (Katow et aZ., 1982; Wessel et aL, 1984). The demonstration of a developmentally regulated increase in affinity for fibronectin thus lends further support to the hypothesis that fibronectin serves as a substrate for morphogenetic movements. We thank G. Cannon for technical assistance in monoclonal antibody production. Laminin was the kind gift of H. Kleinman. We thank M. Truschel, K. Lakoski, G. Wessel, and D. Papaj for critical comments on the manuscript. This work was supported by NIH Grant HD14483 (D.R.M.) and a COCOS Foundation Predoctoral Training Grant in Morphology (R.D.F.). 74 DEVELOPMENTALBIOLOGY REFERENCES BOUCAUT,J. C. and DARRIBERE, T. (1983). Fibronectin in early amphibian embryos. CeU Tissue Res. 234, 135-145. CRITCHLEY,D. R., ENGLAND, M. A., WAKELY, J., and HYNES, R. 0. (1979). Distribution of fibronectin in the ectoderm of gastrulating chick embryos. Nature (Lo&m) 280, 498-500. GIBBONS,J. R., TILNEY, L. G., and PORTER,K. R. (1969). Microtubules in the formation and development of the primary mesenchyme in Arbocia punctulata J. Cell Biol 41, 201-226. GUSTAFSON,T., and KINNANDER, K. (1956). Microaquaria for timelapse cinematographic studies of morphogenesis in swimming larvae and observations on sea urchin gastrulation. Exp. Cell Res. 11,36-51. GUSTAFSON,T., and WOLPERT,L. (1967). Cellular movement and contact in sea urchin morphogenesis. Bid Rev. 42,442-498. HARKEY, M. A., and WHITELEY, A. H. (1981). Isolation, culture, and differentiation of echinoid primary mesenchyme cells. Wilhelm Roux’s Arch. Dev. Biol. 189, 111-122. KATOW, H., and SOLURSH,M. (1979). Ultrastructure of blastocoel material in blastulae and gastrulae of the sea urchin Lytechinus pi&s. J. Exp. Zod 210, 561-567. KATOW, H., YAMADA, K. M., and SOLURSH,M. (1982). Occurrence of fibronectin on the primary mesenchyme cell surface during migration in the sea urchin embryo. mewntiatiun 22, 120-124. LEE, G., HYNES, R. G., and KIRSCHNER,M. (1984). Temporal and spatial regulation of fibronectin in early Xelzopus development. Cell 36, 729-740. MCCLAY, D. R., CANNON,G. W., WESSEL, G. M., FINK, R. D., and MARCHASE,R. B. (1983). Patterns of antigenic expression in early sea urchin development. In “Time, Space, and Pattern in Embryonic Development” (W. R. Jeffery and R. A. Raff, eds.), pp. 157-169. Liss. New York. VOLUME107, 1985 MCCLAY, D. R., and FINK, R. D. (1982). The role of hyalin in early sea urchin development. Dew. Biol 92,285-293. MCCLAY,D. R., and MARCHASE,R. B. (1979). Separation of ectoderm and endoderm from sea urchin pluteus larvae and demonstration of germ layer specific antigens. Dev. Biol 71,289-296. MCCLAY, D. R., WESSEL,G. M., and MARCHASE,R. B. (1981). Intercellular recognition: Quantitation of initial binding events. Proc Nat1 Awd. Sci. USA 78.4975-4979. OKAZAKI, K. (1960). Skeleton formation of sea urchin larvae II. Organic matrix of the spicule. Eminyologia 5, 283-320. OKAZAKI, K. (1965). Skeleton formation of sea urchin larvae. V. Continuous observations of the process of matrix formation. Exp. Cell Res. 40,585-596. OKAZAKI, K. (1975). Spicule formation by isolated micromeres of the sea urchin embryo. Amer. Zd 15,567-581. ROVASIO,R. A., DELOUVEE,A., YAMADA, K. M., TIMPL, R., and THIERY, J. P. 1983).Neural crest cell migration: Requirements for exogenous fibronectin and high cell density. J. Cell Biol 96,462-473. SOLURSH,M., and REVEL,J. P. (1978). A scanning electron microscope study of cell shape and cell appendages in the primitive streak region of the rat and chick embryo. D#wentiatia 11,185-190. SPIEGEL,E., BURGER,M., and SPIEGEL,M. (1986). Fibronectin in the developing sea urchin embryo. J. Cell Biol. 87, 308-313. TOSNEY,K. W. (1978). The early migration of neural crest cells in the trunk region of the avian embryo. An electron microscopic study. Dew. Biol. 62,317-333. UMBREIT,J., and ROSEMAN,S. (1975). A requirement for reversible binding between aggregating embryonic cells before stable adhesion. J. Biol. Chem. 250,9360-9368. WESSEL,G. M., MARCHASE,R. B., and M&LAY, D. R. (1984). Ontogeny of the basal lamina in the sea urchin embryo. Dew. Biol. 103, 235245.