- Email: [email protected]
Inhibition of cell migration in sea urchin embryos by β-d -xyloside
DEVELOPMENTAL
BIOLOGY
118,325-332
(1986)
Inhibition of Cell Migration in Sea Urchin Embryos by ,8-D-Xyloside MICHAEL *Department
of Biology,
SOLURSH,*
University
SUSAN LENOCH MITCHELL,*
AND HIDEKI
of Iowa, Iowa City, Iowa 52242, and TBiolcgy Laboratory, Received June 12, 1985; accepted in revised
fwm July
KATOW~
Rikkyo
University,
Tokyo 171, Japan
7, 1986
This investigation examines the effect of exogenous xylosides on primary mesenchyme cell behavior in Sr~titrotus embryos. In confirmation of studies in some other species the addition of 2 mM pnitrophenyl-@-D-xylopyranoside blocks the migration but not the initial ingression of primary mesenchyme cells. The blastocoel matrix of treated embryos appears deficient in a 15- to 30-nm-diameter granular component that is observed extensively on the basal lamina and on filopodia of migrating primary mesenchyme cells in untreated embryos. Other blastocoel components appear unaffected by ultrastructural criteria. The incorporation of 35SOq-per embryo into ethanol precipitates of isolated blastocoel matrices was reduced significantly after xyloside treatment but the distribution of “SO,“- after polyacrylamide gel electrophoresis or the glycosaminoglycan composition was unaffected. Chromatography on Sepharose CL-2B demonstrates a reduction in size of sulfated components of the blastocoel. While over 60% of the “S-labeled material from the blastocoel of normal mesenchyme blastulae is voided from a Sepharose CL-2B column run in a dissociative solvent, only 10% from xyloside treated embryos is voided. Instead, there is a large included peak with K., of 0.33. This material is acid soluble but cetylpyridinium chloride precipitahle. It apparently consists largely of free glycosaminoglycan chains. Based on analysis of chondroitinase ABC digestion products this material consists of 41% chondroitin-6-sulfate and 53% dermatan sulfate. These results are consistent with a role in cell migration for intact chondroitin sulfate/dermatan sulfate proteoglycans in the sea urchin blastocoel matrix. 0 1986 Academic Press, Inc. pwparatus
Affected glycosaminoglycans are normally linked by xylose to specific serine residues in the core protein, including chondroitin sulfate (Galligani et aZ., 1975; Fukunaga et al, 1975; Gibson and Segin, 1977) and heparan sulfate (Hart and Lennarz, 1978). Dermatan sulfate often has a similar xyloside linkage (Akiyama and Seno, 1981), but apparently not in some sea urchin species (Oguri and Yamagata, 1978). Treatment of sea urchin embryos with /3-xylosides has an effect on primary mesenchyme cell migration similar to that of sulfate deprivation. Treatment depresses proteoglycan synthesis and permits ingression, but blocks migration (Kinoshita and Sagai, 1979; Akasaka et al., 1980). The present report extends previous studies by correlating the effects of xyloside treatment on the ultrastructure of the blastocoel matrix with those on sulfated proteoglycans extracted from the blastocoel matrix.
INTRODUCTION
The primary mesenchyme cells of the sea urchin embryo have served as a useful model system in which to study embryonic cell migration in situ (see Solursh, 1986). These cells ingress from the vegetal plate and after a lag period migrate using the basal lamina that surrounds the blastocoel as the substratum (Katow and Solursh, 1980; 1981). Sulfated proteoglycans are thought to play a role in the migration of these cells. With sulfate deprivation of Lytechinus pictus embryos, the primary mesenchyme cells ingress and move around the vegetal plate, but fail to migrate up the blastocoel wall (Karp and Solursh, 1974). Time-lapse cinemagraphic studies indicate that cell processes fail to form stable attachments to the substratum (Katow and Solursh, 1981). Sulfate deprivation results in the reduction in a 15 to 30-nm-diameter granular component in the blastocoel and basal lamina (Katow and Solursh, 1979). The major sulfated macromolecules in the blastocoel have been shown to consist of chondroitin-6-sulfate, dermatan sulfate, and heparan sulfate proteoglycans (Solursh and Katow, 1982). A more specific treatment which interferes with the synthesis of normal proteoglycans is the addition of exogenous P-xylosides. Several /3-xyloside derivatives substitute in vivo for proteoglycan core proteins in the initiation of synthesis of some glycosaminoglycans. Treatment results in the synthesis of free glycosaminoglycan chains and glycosaminoglycan-depleted core proteins.
MATERIALS
AND METHODS
Embryo Culture The eggs of the sea urchin Strmgylocentrotus pqwuratus (Pacific Biomarine, Venice, Calif.) were collected by injection of 0.5 MKCI. Inseminated eggs were incubated in either artificial seawater (Dawson et d, 1969), sulfatefree seawater (Karp and Solursh, 1974), or seawater containing 2 or 4 mM of p-nitrophenyl-P-D-xylopyranoside, p-nitrophenyl-a-D-xylopyranoside (Koch-Light 325
0012-1606/86 $3.00 Copyright All rights
0 1986 by Academic Press, Inc. of reproduction in any form reserved.
326
DEVELOPMENTAL BIOLOGY VOLUME118,1986
Laboratories), p-nitrophenyl galactopyranoside (Sigma), al. (1968). The digestion products were separated by deD-gdaCtOSe (Sigma), or D-xylose (Pfanstiehl) at 16°C scending paper chromatography. Nitrous acid sensitive until the mesenchyme blastula stage. Radiolabeling was and resistant components were separated by chromaperformed as described earlier (Solursh and Katow, tography in a Sephadex G-50 column. 1982). Hatched blastula stage embryos (2-4 X lo4 emRadiolabeled blastocoel matrix was analyzed by colbryos/ml) were incubated at 16”/C for 2 hr until the umn chromatography. In some cases,blastocoel material mesenchyme blastula stage in seawater containing 100 was pretreated for 1 hr at 37°C with 10 units/ml chon&i/ml ?S04 (43 Ci/mg) or 6-20 pCi/ml L-[6-3H]fucose droitinase ABC, as described earlier (Solursh and Katow, (lo-15 Ci/mmole) (New England Nuclear Corp.). The 1982). Treated or untreated samples were diluted to 1 embryos were washed in cold seawater and the blastocoel ml with 4 M guanidine HCl in 0.05 M Na acetate (pH matrix was isolated at 4°C as described below. 5.8) and run on a 0.8 X 24-cm Sephadex G-50 column. The voided fractions were pooled, concentrated, and loaded on a 115 X 1.3-cm analytical column of Sepharose Microscopic Examination CL-2B. One-milliliter fractions were collected and For ultrastructural studies, specimens were prepared counted in Aquasol II in a Beckman LS 200 liquid scinas described previously (Katow and Solursh, 1981). tillation counter. Briefly, embryos were fixed in 2.5% glutaraldehyde in The pooled Sepharose CL-2B fractions from xyloside 0.2 Mphosphate buffer (pH 7.4) for 1 hr at 4”C, postfixed treated embryos which were included were subjected to in 1% of Os04 in 0.2 M phosphate buffer for 1 hr at 4°C further analysis. The percentage of counts precipitable and embedded in Epon. Thin sections were stained with in cold 10% trichloroacetic acid in the presence of 10% uranyl acetate and lead citrate (Reynolds, 1963), and serum as carrier, was determined. The label present in examined with a Philips EM 300 TEM. Aliquots of em- glycosaminoglycans was determined by the addition of bryos were critical point dried (Anderson, 1951) after cetylpyridinium chloride to 0.5% in the presence of 0.5 dehydration in ethanol. The dried specimens were mg/ml chondroitin sulfate (Sigma) as carrier (Karp and broken open with a mounted hair, coated with gold Solursh, 1974). The precipitates were washed once in palladium, and observed with a JEOL JSM-35C SEM cold water and dissolved in methanol, and then counted. at 14 kV. Pooled fractions were also treated for 3 hr at 37°C with Tris-enriched buffer, 10 units/ml chondroitinase AC or Isolaticm of Blastocoel Matrix chondroitinase ABC (Miles Laboratory) prior to chroThe extracellular materials, including the blastocoel matography on a 1 X 25-cm Sephadex G-75 column in material and the basal lamina, were isolated from mes- 0.1 M NH4 acetate. One milliliter fractions were collected enchyme blastulae by a modification of the method of and counted in the scintillation counter. Harkey and Whiteley (1980), as described previously RESULTS (Solursh and Katow, 1982). Briefly, the embryos were washed in Ca2+, Mgz+-free seawater containing 100 M Eflect of Xyloside on Development EDTA and dispersed in a solution containing 1 Mglycine Embryos cultured in 2 or 4 mM p-nitrophenyl-P-Dand 100 uM EDTA (pH 8). After 15 min the dissociated xylopyranoside (B-xyloside) developed normally until the cells were centrifuged and the supernatant, containing mesenchyme blastula stage. The primary mesenchyme the cell-free matrix, was collected. Three volumes of cold cells ingressed but then failed to migrate (Fig. l), al95% ethanol were added and the samples were stored at though they were able to form cell processes [based on -20°C until further use. time-lapse cinematography (not shown)]. Further development was blocked at this stage, although the emCharacterization of Radiolabeled Blastocoel Ma&ix bryos were motile and appeared healthy for at least anBlastocoel matrix was recovered by centrifugation af- other day. At these concentrations of either a- or fiter ethanol precipitation. Analysis on one-dimensional xyloside, mesenchyme cell migration was blocked in at 7.5% polyacrylamide slab gels was carried out according least 91% of the embryos, based on counts of over 1.5 to the method of Laemmli (1970), as described previously X lo3 embryos. This contrasts with extensive mesen(Solursh and Katow, 1982). Characterization of sulfate- thyme migration in over 93% of untreated control emlabeled glycosaminoglycans was also carried out as de- bryos. At lower concentrations (e.g.,
SOLURSH,MITCHELL, AND KATOW
Cell Migration
in Sea Urchin Embryos
327
a FIG. 1. Phase-contrast micrographs of the normal mesenchyme blastula (a) and an embryo treated with 2 mM @-D-xyloside, in which primary mesenchyme cell migration does not occur (b). The ingressed primary mesenchyme cells are migrating along the inner blastocoel wall toward the animal pole (A) in normal embryos (arrows in (a)) while those in the xyloside-treated embryo have ingressed but remain piled up at the vegetal plate (V) (arrows in (b)) X1100.
side. In the best case, only 55% of the embryos reared in sulfate-free seawater failed to undergo normal primary mesenchyme cell migration. In contrast to these treatments, culture of embryos in 2 or 4 mM nitrophenyl galactopyranoside, D-galactose, or D-xylose had no observable effect on development including primary mesenchyme cell migration. Morphology of the Blastocoel 2Matm.x Normal mesenchyme blastulae. By transmission electron microscopy, the blastocoel material appears as a complex of 15- to 30-nm granules and ca. 30-rim-diameter fibers (Fig. 2). The granules are often clustered, among which short, thin fibers about 4 nm in diameter are observed (Fig. 2, inset). The basal lamina also contains 15to 30-nm-diameter granules. These are localized on the blastocoel side of the basal lamina, while a fibrous material about 6 nm in diameter forms a network on the ectodermal side of the basal lamina (Fig. 2b). These fibers in the basal lamina appear morphologically different from the 4-nm fibers in the blastocoel, while the granules in the two locations are not morphologically distinguishable. The granules are also observed on the surfaces of cell processes extending from primary mesenchyme cells (Fig. 3), whether or not they are near the basal lamina. Xyloside-treated mesenchyme blastulae. Treatment with xyloside has a morphological effect on the blastocoel matrix. By transmission electron microscopy, the 30-nm granules are much less numerous than in normal embryos (Fig. 4) and clusters of the granules are seldom observed (Fig. 4a). On the other hand, the 4-nm-diameter fibers are often seen. The 30-nm granules are also less
numerous in the basal lamina, although the fibrous component is conspicuous (Fig. 4b). Effect of Xyloside on Sulfated Proteoglycans in the Blastocoel As considered, in many systems fi-xyloside stimulates the synthesis of some glycosaminogylcans (GAG). However, these are not protein bound but are free GAG chains. In contrast, treatment of S. purpuratus with 2 mM P-D-xyloside depresses the incorporation of 35SOzinto ethanol precipitable, nondialyzable material in the blastocoel matrix, where treated preparations ranged from 88 to 60% of control embryos expressed as CPM per embryo. The proportions of the various types of sulfated GAG present in Pronase digests were indistinguishable between normal controls and xyloside-treated embryos. As reported previously for normal embryos (Solursh and Katow, 1982), about 16% of the 35SOi- is found in chondroitin-6-S04, 37% in dermatan sulfate, 12% in heparan sulfate, 20% in glycopeptides, and the remainder still uncharacterized (Table I). Autofluorograms after sodium dodecyl sulfate-polyacrylamide gel electrophoresis of the sulfate-labeled blastocoel matrix from normal and xyloside-treated embryos on 7.5% gels are qualitatively indistinguishable from each other (not shown). Since proteoglycans are the major sulfated component in the blastocoel (Solursh and Katow, 1982) and these are likely targets of P-D-xyloside, the proteoglycans are examined in more detail. The size of the sulfate-labeled blastocoel components is analyzed by chromatography on a Sepharose CL-2B column using a dissociative solvent. As reported previously (Solursh and Katow, 1982),
328
DEVELOPMENTALBIOLOGY
VOLUME118, 1986
FIG. 2. TEM of normal mesenchyme blastula. The blastocoel material is composed of thin, long fibers (arrows in (a)) and granules, which are ca. 15-30 nm in diameter (large arrows in inset). These granules are associated with thin, short fibers (small arrows in inset). The basal lamina, on the other hand, is also composed of a fibrous material (thin arrows in (b)), which is morphologically different from the other fibrous structures in the blastocoel, and the 15- to 30-nm-diameter granules are attached to the blastocoel side of the basal lamina so that the basal lamina is polarized. Open arrows in (b) indicate the thin, short fibers of the blastocoel material. (a) X63,000. Inset, X145,000. (b) X128,000.
the majority of the label from normal embryos is eluted in the void volume (at least 60%, Fig. 5). Much of this label is sensitive to chondroitinase ABC pretreatment (Fig. 5), as reported earlier (Solursh and Katow, 1982). In contrast, only 10% of the % from the blastocoel matrix of xyloside-treated embryos was in the voided fractions. The majority of the label was included with about 44% having a Kay of 0.33. In this case too, the majority of the label is sensitive to chondroitinase ABC pretreatment, with a minor resistant component remaining in the voided fractions. These results suggest that while the overall GAG composition of blastocoel matrix is
similar in normal and treated embryos, its organization in proteoglycans is drastically altered. Based on results in other systems, one possibility is that xyloside treatment causes the formation of GAGdeficient proteoglycans and free GAG chains. The latter would be expected in the included fractions of the CL2B column. This possibility was examined by further analysis of the included fractions in Fig. 5 (fractions 65-85). While the majority of the counts are precipitated with cetylpyridinium chloride, only 10% of these counts were precipitated in cold TCA, suggesting that this material is largely in the form of free GAG chains. Based
SOLURSH, MITCHELL, AND KATOW
Cell Migration
in Sea Urchin
Embryos
329
on sensitivity to chondroitinase AC and ABC (Fig. 6), this material appears to be a mixture of chondroitin sulfate [chondroitin-6-SO1 based on earlier studies (Solursh and Katow, 1982)] and dermatan sulfate, in a ratio of 1:1.4, respectively. A sulfated fucogalactan-protein conjugate has been described in the extracellular matrix of sea urchin embryos (Akasaka and Terayama, 1982; 1983; 1934). Blastocoel material, from normal and xyloside-treated embryos, prepared after metabolic labeling with [3H]fucose was analyzed by chromatography on a Sephacose CL2B column (not shown). In this case, the elution patterns of samples from normal and xyloside-treated embryos were similar. About 30-45s of the recovered label was in the voided fractions and 45-58s was included. These results suggest that fucose rich components in the blastocoel are relatively resistant to xyloside treatment. DISCUSSION FIG. 3. TEM of primary mesenchyme cell process extended onto the basal lamina (BL) which is associated with 15- to 30-rim-diameter granules (arrows).
The morphological features of the blastocoel matrix in the mesenchyme blastula stage of S. purpuratus are
FIG. 4. TEM of mesenchyme blastula treated with 2 mM xyloside. The 15- to 30-nm-diameter granules are drastically reduced from both the blastocoel material (large arrows in (a)) and the basal lamina (large arrow in (b)) so that the short, thin fibers of the blastocoel material (small arrows in (a)) and the fibrous material of the basal lamina (small arrows in (b)) become more apparent than in the normal embryo (Fig. 2). These fibers appear intact morphologically. (a) X120,000 (b) X128,000.
330
DEVELOPMENTAL BIOLOGY
VOLUME 118, 1986
TABLE 1 S&-LABELED MATERIAL&JIN PRONASE DIGESTS OF BLASTOCOELMATRIX FROM NORMAL AND XYLOSIDE-TREATED MESENCHYME BLASTULAE Normal (%I Sulfated glycoprotein” Heparan sulfate* Chondroitin-6-SO,” Dermatan Sold Uncharacterized
20 12 16 37 15
Xyloside-treated (%) 23 12 16 36 13
Note. These results are typical of three separate trials. ‘The percentage of the total counts in the included fractions on a Sephadex G-50 column. bThe percentages of the total counts that are sensitive to nitrous acid treatment prior to chromatography on Sephadex G-50. “The percentages of the total counts that are sensitive to chondroitinase AC and cochromatograph with 2-acetamide-2-deoxy-3-0(fi-D-gluco-4-enepyranosyluronic acid)-6-sulfo-D-galactose standard. dThe percentages of the total counts that are sensitive to chondroitinase ABC and cochromatograph with either 2-acetamide-e-deoxy3-G-(fl-D-gluco-4-enepyranosyluronic acid)-D-galactose 4,gdisulfate or -4 or -6-sulfo-D-galactose minus the chondroitinase AC-sensitive counts.
generally like those described earlier in L pict~ (Katow and Solursh, 1979). The 15 to 30-nm granules are a prominent component and are present in the blastocoel on the blastocoel side of the basal lamina and on filopodia of migrating primary mesenchyme cells. They have the localization that one would expect for a component that is involved in the attachment of cells to the substratum. Notably, these granules are present at sites of cell attachment to the basal lamina (Katow and Solursh, 1981). Unlike S. pwpuratus, the thin fibrous component of L. pictus is not observed until the onset of gastrulation. Based on the morphology of these granules and their affinity for ruthenium red (Katow and Amemiya, 1986), they are likely to be proteoglycan granules (Hay, 1978). In addition to the association of the granules with sites of mesenchyme cell attachment, treatments that interfere with their accumulation inhibit cell migration. Sulfate deprivation of L pictus embryos permits primary mesenchyme cell ingression but prevents their migration (Karp and Solursh, 1974). The cells appear to be unable to form stable cell attachments to the basal lamina (Katow and Solursh, 1981). This treatment results in a paucity of the 15- to 30-nm granules on the basal lamina and in the blastocoel (Katow and Solursh, 1979). Sulfate deprivation of Clypeaetcrja~icus embryos inhibits the synthesis of dermatan sulfate (Yamaguchi and Kinoshita, 1985). Xyloside treatment, which is presumably a more specific inhibitor of the synthesis of normal proteoglycans than sulfate deprivation, has a similar effect. Primary mesenchyme cells ingress but fail to migrate (Kinoshita
"0
30
40
SO
60
70
n "
60
60
100
FIG. 5. S&-labeled hlastocoel matrix analyzed on a Sepharose CL2B column. S%-labeled blastocoel matrix from normal (0) or xylosidetreated (0) mesenchyme blastulae was analyzed on an analytical Sepharose CL-2B column with 4 Mguanidine HCl as solvent. At least 60% of the S% from normal embryos was in the void volume (V,). On the other hand, only 10% of the Ss6 from xyloside-treated embryos was voided. The remainder was spread throughout the included region with about 44% eluting with a K., of 0.33. If the samples were treated with chondroitinase ABC and passed through Sephadex G-50 prior to chromatography on Sepharose CL-LB, only a small percentage of the original counts from normal (0) and xyloside-treated (m) embryos is recovered. This material is largely restricted to the voided fractions. These results are typical of two trials.
ML
FIG. 6. Chondroitinase sensitivity of included fractions of blastocoel matrix from xyloside-treated embryos. S”-labeled blastocoel matrix from xyloside-treated embryos was analyzed on a dissociative CL-2B column and the major included fractions (fractions 65-85 in Fig. 5) were pooled, dialyzed, lyophilized, and run on a Sephadex G-75 column without prior treatment (0) or with prior treatment with either chondroitinase ABC (m) or chondroitinase AC (A). While 90% of the counts in the untreated sample are voided, 94% of the counts were in the Vt fractions after treatment with chondroitinase ABC. After treatment with chondroitinase AC, 38% of the counts are still voided with the remainder included or in the V,. Recovery was greater than 80% in all cases.
SOLURSH, MITCHELL, AND KATOW
Cell Migration
and Saiga, 1979; Akasaka et al, 1980; and Fig. 1). The present study shows that treatment results in a selective reduction of the 15 to 30-nm-diameter granules. Kinoshita and Saiga (1979) found that the inhibitory effects were observed only with P-xyloside and not with the (Yisomer, in contrast to the present study where both isomers are effective. This difference could be the result of differences in the ability to convert the cyisomer to the active /3 form. The specificity of xylosides is suggested by the absence of any detectible effect by D-galactose and fi-D-galactopyranoside. Xylose is also ineffective and is much less active in initiating the synthesis of glycosaminoglycans in other systems (Galligani et ak, 1975; Gibson and Segin, 1977). Other species differences include the relative resistance of primary mesenchyme cell migration in S. purpuratus to sulfate deprivation compared to L. pi&m and the relative resistance of primary mesenchyme cell migration in L. pictus to xylosides (unpublished observation). The biochemical comparisons of %S-labeled blastocoel matrix from untreated and xyloside-treated embryos are consistent with a role of a chondroitin sulfate/dermatan sulfate proteoglycan in primary mesenchyme cell migration. While the proportions of the various types of glycosaminoglycans in the blastocoel remain unaffected by xyloside treatment, with over half of the radiolabeled sulfated material being in the form of dermatan sulfate and chondroitin-6-sulfate, the total incorporation is reduced significantly. This later result agrees with the earlier report of Kinoshita and Saiga (1979). It should be recognized that only blastocoel material is analyzed here and some free glycosaminoglycan chains formed in the presence of xyloside could be present in other embryo fractions. The sulfated proteoglycans considered in this report are clearly heterogeneous and require further characterization. Even after xyloside treatment there remains additional chondroitinase ABC-sensitive label. Xyloside treatment appears to reduce the size of the major sulfated proteoglycans in the blastocoel matrix. Over 60% of the label from untreated embryos is voided from a Sepharose CL-ZB column and appears to be larger than 40 X lo6 Da. Much of the label is sensitive to chondroitinase ABC, as reported earlier (Solursh and Katow, 1982). Only 10% of the label is voided when material from xyloside treated embryos is examined and a large proportion of the 35SOf- is included with a K,, of 0.33. Most of this later material is sensitive to chondroitinase ABC and appears to consist of free glycosaminoglycan chains, as one would expect after xyloside treatment. The relative increased proportion of chondroitin sulfate to dermatan sulfate (1:1.4) in the included fractions compared to that in the whole extract (1:2.3) suggests heterogeneity in the dermatan sulfate containing proteoglycans. While dermatan sulfate is normally xyloside
in Sea Urchin Embryos
331
linked (Akiyama and Seno, 1981), this might not be the case for all sea urchin dermatan sulfate (Oguri and Yamagata, 1978). The present results are consistent with the existence of xyloside linked dermatan sulfate and chondroitin sulfate-rich proteoglycans as well as dermatan sulfate proteoglycans that do not contain an Oglycosidic bond. While this possibility remains to be investigated, xyloside linked chondroitin sulfatejdermatan sulfate-rich proteoglycans may be specifically involved in cell migration. It is not yet known if a single proteoglycan containing both chondroitin sulfate and dermatan sulfate mediates mesenchyme cell attachment during migration. This work was supported
by NIH Grant HD16549. REFERENCES
AKASAKA, K., AMENIYA, S., and TERAYAMA, H. (1980). Scanning electron microscopical study of the inside of sea urchin embryos (Pseudo-
eentrotus depress+&. Exp. Cell Res. 129,1-13. K., and TERAYAMA, H. (1982). Sulfated fucogalactan-protein conjugate present in the EDTA-extract from Anthocidaris embryos (mid-gastrula). J. Fat. Sci Univ. Tokgo. 15,181-189. AKASAKA, K., and TERAYAMA, H. (1983). Sulfated glycan present in the EDTA extract of Hemicentrotus embryo (mid-gastrula). Exp. Cell. Res, 146,177-185. AKASAKA, K., and TERAYAMA, H. (1984). A proteoglycan fraction isolated from the EDTA extract of sea urchin (Hemicentrotus pulcherrimus) gastrulae stimulates reaggeration of dissociated embryonic cells. Exp. Cell Res. 150, 226-233. AKIYAMA, F., and SENO, N. (1981). Linkage regions between dermatan polysulfates and peptides. B&him. Biophys. Acta 674,280-296. ANDERSON, T. (1951). Technique for preservation of 3-dimensional structure in preparing specimens for the E. M. Trans. N. Y. Acad sci 13,130-134. DAWSON, R. M. C., ELLIOTT, W. H., and JONES, K. M. (1969). “Data for Biochemical Research.” p. 508. Oxford Univ. Press, London/New York. FUKUNAGA, Y., SOBUE, M., SUZUKA, N., KUSHIDA, H., and SUZUKI, S. (1975). Synthesis of a fluorogenic mucopolysaccharide by ehondrocytes in cell culture with I-methylumbelliferyl @D-xyloside. B&him. Biophys. Acta 381,443-447. GALLIGANI, L., HOPWOOD,J., SCHWARTZ, N. B., and DORFMAN, A. (1975). Stimulation of synthesis of free chondroitin sulfate chains by O-Dxyloside in cultured cells. J. Biol. Chem. 250,5400-5406. GIBSON, K. D., and SEGEN, B. J. (1977). The mode of action of 4-methylumbelliferyl @-D-xyloside on the synthesis of chondroitin sulphate in embryonic-chicken sternum. Biochem J. 168,65-79. HARKEY, M. A., and WHITELEY, A. H. (1980). Isolation, culture, and differentiation of echinoid primary mesenchyme cells. Rwx’s Arch. 189.111-122. HART, G. W., and LENNARZ, W. J. (1978). Effects of tunicamycin on the biosynthesis of glycosaminoglycans by embryonic chick cornea. J. AKASAKA,
Biol. Chem. 253,5’795-5801. E. D. (1978). Fine structure of embryonic matrices and their relation to the cell surface in ruthenium red-fixed tissues. Growth 42,399-423. KARP, G. C., and SOLURSH, M. (1974). Acid mycopolysaccharide metabolism, the cell surface, and primary mesenchyme cell activity in the sea urchin embryo. Deu. Biol. 41.110-123. KATOW, H., and AMEMIYA, S. (1986). Behavior of primary mesenchyme cells in situ associated with ultrastructural alterations of the blasHAY,
332
DEVELOPMENTALBIOLOGY
tocoelic material in the sea urchin, Anthocidaris crastipi~~~. Dev. Growth D@Y. 28.31-42. KATOW,H., and SOLURSH,M. (1979). Ultrastructure of blastocoel material in blastulae and gastrulae of the sea urchin Lytechinus pictus. J. Exp. ZooL 210,561-567. KATOW, H., and SOLURSH,M. (1980). Ultrastructure of primary mesenchyme ingression in the sea urchin Lytechinus pi&us. J. Exp. Zool. 213,231-246. KATOW, H., and SOLURSH,M. (1981). Ultrastructural and time-lapse studies of primary mesenchyme cell behavior in normal and sulfatedeprived sea urchin embryos. Exp. Cell Res. 136,233-245. KINOSHITA, S., and SAIGA, H. (1979). The role of proteoglycan in the development of sea urchins. Exp. Cell Res. 123,229-236. LAEMMLI, U. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 227,680685.
VOLUME118, 1986
OGURI,K., and YAMAGATA,T. (1978). Appearance of a proteoglycan in developing sea urchin embryos. Biochim Biophys. Actu 541.385-393. REYNOLDS, E. S. (1963).The use of lead citrate at high pH as an electronopaque stain in electron microscopy. J. CeUBiol. 17,208-212. SAITO, H., YAMAGATA,T., and SUZUKI, S. (1968). Enzymatic methods for the determination of small quantities of isomeric chondroitin sulfate. J. BioL Chem 243,1536-X42. SOLURSH,M. (1986).Migration of sea urchin primary mesenchyme cells. In “Developmental Biology: A Comprehensive Synthesis.” Vol. 2, pp. 391-431. Plenum, New York. SOLURSH,M., and KATOW,H. (1982). Initial characterization of sulfated macromolecules in the blastocoels of mesenchyme blastulae of Stmng&xentrotus purpuratus and Lytechinus pi&us. De-v.Biol 94, 326-336. YAMAGUCHI,M., and KINOSHITA, S. (1985). Polysaccharides sulfated at the time of gastrulation in embryos of the sea urchin Clypeaster japonicw. Exp. Cd Res. 159,353-365.
BIOLOGY
118,325-332
(1986)
Inhibition of Cell Migration in Sea Urchin Embryos by ,8-D-Xyloside MICHAEL *Department
of Biology,
SOLURSH,*
University
SUSAN LENOCH MITCHELL,*
AND HIDEKI
of Iowa, Iowa City, Iowa 52242, and TBiolcgy Laboratory, Received June 12, 1985; accepted in revised
fwm July
KATOW~
Rikkyo
University,
Tokyo 171, Japan
7, 1986
This investigation examines the effect of exogenous xylosides on primary mesenchyme cell behavior in Sr~titrotus embryos. In confirmation of studies in some other species the addition of 2 mM pnitrophenyl-@-D-xylopyranoside blocks the migration but not the initial ingression of primary mesenchyme cells. The blastocoel matrix of treated embryos appears deficient in a 15- to 30-nm-diameter granular component that is observed extensively on the basal lamina and on filopodia of migrating primary mesenchyme cells in untreated embryos. Other blastocoel components appear unaffected by ultrastructural criteria. The incorporation of 35SOq-per embryo into ethanol precipitates of isolated blastocoel matrices was reduced significantly after xyloside treatment but the distribution of “SO,“- after polyacrylamide gel electrophoresis or the glycosaminoglycan composition was unaffected. Chromatography on Sepharose CL-2B demonstrates a reduction in size of sulfated components of the blastocoel. While over 60% of the “S-labeled material from the blastocoel of normal mesenchyme blastulae is voided from a Sepharose CL-2B column run in a dissociative solvent, only 10% from xyloside treated embryos is voided. Instead, there is a large included peak with K., of 0.33. This material is acid soluble but cetylpyridinium chloride precipitahle. It apparently consists largely of free glycosaminoglycan chains. Based on analysis of chondroitinase ABC digestion products this material consists of 41% chondroitin-6-sulfate and 53% dermatan sulfate. These results are consistent with a role in cell migration for intact chondroitin sulfate/dermatan sulfate proteoglycans in the sea urchin blastocoel matrix. 0 1986 Academic Press, Inc. pwparatus
Affected glycosaminoglycans are normally linked by xylose to specific serine residues in the core protein, including chondroitin sulfate (Galligani et aZ., 1975; Fukunaga et al, 1975; Gibson and Segin, 1977) and heparan sulfate (Hart and Lennarz, 1978). Dermatan sulfate often has a similar xyloside linkage (Akiyama and Seno, 1981), but apparently not in some sea urchin species (Oguri and Yamagata, 1978). Treatment of sea urchin embryos with /3-xylosides has an effect on primary mesenchyme cell migration similar to that of sulfate deprivation. Treatment depresses proteoglycan synthesis and permits ingression, but blocks migration (Kinoshita and Sagai, 1979; Akasaka et al., 1980). The present report extends previous studies by correlating the effects of xyloside treatment on the ultrastructure of the blastocoel matrix with those on sulfated proteoglycans extracted from the blastocoel matrix.
INTRODUCTION
The primary mesenchyme cells of the sea urchin embryo have served as a useful model system in which to study embryonic cell migration in situ (see Solursh, 1986). These cells ingress from the vegetal plate and after a lag period migrate using the basal lamina that surrounds the blastocoel as the substratum (Katow and Solursh, 1980; 1981). Sulfated proteoglycans are thought to play a role in the migration of these cells. With sulfate deprivation of Lytechinus pictus embryos, the primary mesenchyme cells ingress and move around the vegetal plate, but fail to migrate up the blastocoel wall (Karp and Solursh, 1974). Time-lapse cinemagraphic studies indicate that cell processes fail to form stable attachments to the substratum (Katow and Solursh, 1981). Sulfate deprivation results in the reduction in a 15 to 30-nm-diameter granular component in the blastocoel and basal lamina (Katow and Solursh, 1979). The major sulfated macromolecules in the blastocoel have been shown to consist of chondroitin-6-sulfate, dermatan sulfate, and heparan sulfate proteoglycans (Solursh and Katow, 1982). A more specific treatment which interferes with the synthesis of normal proteoglycans is the addition of exogenous P-xylosides. Several /3-xyloside derivatives substitute in vivo for proteoglycan core proteins in the initiation of synthesis of some glycosaminoglycans. Treatment results in the synthesis of free glycosaminoglycan chains and glycosaminoglycan-depleted core proteins.
MATERIALS
AND METHODS
Embryo Culture The eggs of the sea urchin Strmgylocentrotus pqwuratus (Pacific Biomarine, Venice, Calif.) were collected by injection of 0.5 MKCI. Inseminated eggs were incubated in either artificial seawater (Dawson et d, 1969), sulfatefree seawater (Karp and Solursh, 1974), or seawater containing 2 or 4 mM of p-nitrophenyl-P-D-xylopyranoside, p-nitrophenyl-a-D-xylopyranoside (Koch-Light 325
0012-1606/86 $3.00 Copyright All rights
0 1986 by Academic Press, Inc. of reproduction in any form reserved.
326
DEVELOPMENTAL BIOLOGY VOLUME118,1986
Laboratories), p-nitrophenyl galactopyranoside (Sigma), al. (1968). The digestion products were separated by deD-gdaCtOSe (Sigma), or D-xylose (Pfanstiehl) at 16°C scending paper chromatography. Nitrous acid sensitive until the mesenchyme blastula stage. Radiolabeling was and resistant components were separated by chromaperformed as described earlier (Solursh and Katow, tography in a Sephadex G-50 column. 1982). Hatched blastula stage embryos (2-4 X lo4 emRadiolabeled blastocoel matrix was analyzed by colbryos/ml) were incubated at 16”/C for 2 hr until the umn chromatography. In some cases,blastocoel material mesenchyme blastula stage in seawater containing 100 was pretreated for 1 hr at 37°C with 10 units/ml chon&i/ml ?S04 (43 Ci/mg) or 6-20 pCi/ml L-[6-3H]fucose droitinase ABC, as described earlier (Solursh and Katow, (lo-15 Ci/mmole) (New England Nuclear Corp.). The 1982). Treated or untreated samples were diluted to 1 embryos were washed in cold seawater and the blastocoel ml with 4 M guanidine HCl in 0.05 M Na acetate (pH matrix was isolated at 4°C as described below. 5.8) and run on a 0.8 X 24-cm Sephadex G-50 column. The voided fractions were pooled, concentrated, and loaded on a 115 X 1.3-cm analytical column of Sepharose Microscopic Examination CL-2B. One-milliliter fractions were collected and For ultrastructural studies, specimens were prepared counted in Aquasol II in a Beckman LS 200 liquid scinas described previously (Katow and Solursh, 1981). tillation counter. Briefly, embryos were fixed in 2.5% glutaraldehyde in The pooled Sepharose CL-2B fractions from xyloside 0.2 Mphosphate buffer (pH 7.4) for 1 hr at 4”C, postfixed treated embryos which were included were subjected to in 1% of Os04 in 0.2 M phosphate buffer for 1 hr at 4°C further analysis. The percentage of counts precipitable and embedded in Epon. Thin sections were stained with in cold 10% trichloroacetic acid in the presence of 10% uranyl acetate and lead citrate (Reynolds, 1963), and serum as carrier, was determined. The label present in examined with a Philips EM 300 TEM. Aliquots of em- glycosaminoglycans was determined by the addition of bryos were critical point dried (Anderson, 1951) after cetylpyridinium chloride to 0.5% in the presence of 0.5 dehydration in ethanol. The dried specimens were mg/ml chondroitin sulfate (Sigma) as carrier (Karp and broken open with a mounted hair, coated with gold Solursh, 1974). The precipitates were washed once in palladium, and observed with a JEOL JSM-35C SEM cold water and dissolved in methanol, and then counted. at 14 kV. Pooled fractions were also treated for 3 hr at 37°C with Tris-enriched buffer, 10 units/ml chondroitinase AC or Isolaticm of Blastocoel Matrix chondroitinase ABC (Miles Laboratory) prior to chroThe extracellular materials, including the blastocoel matography on a 1 X 25-cm Sephadex G-75 column in material and the basal lamina, were isolated from mes- 0.1 M NH4 acetate. One milliliter fractions were collected enchyme blastulae by a modification of the method of and counted in the scintillation counter. Harkey and Whiteley (1980), as described previously RESULTS (Solursh and Katow, 1982). Briefly, the embryos were washed in Ca2+, Mgz+-free seawater containing 100 M Eflect of Xyloside on Development EDTA and dispersed in a solution containing 1 Mglycine Embryos cultured in 2 or 4 mM p-nitrophenyl-P-Dand 100 uM EDTA (pH 8). After 15 min the dissociated xylopyranoside (B-xyloside) developed normally until the cells were centrifuged and the supernatant, containing mesenchyme blastula stage. The primary mesenchyme the cell-free matrix, was collected. Three volumes of cold cells ingressed but then failed to migrate (Fig. l), al95% ethanol were added and the samples were stored at though they were able to form cell processes [based on -20°C until further use. time-lapse cinematography (not shown)]. Further development was blocked at this stage, although the emCharacterization of Radiolabeled Blastocoel Ma&ix bryos were motile and appeared healthy for at least anBlastocoel matrix was recovered by centrifugation af- other day. At these concentrations of either a- or fiter ethanol precipitation. Analysis on one-dimensional xyloside, mesenchyme cell migration was blocked in at 7.5% polyacrylamide slab gels was carried out according least 91% of the embryos, based on counts of over 1.5 to the method of Laemmli (1970), as described previously X lo3 embryos. This contrasts with extensive mesen(Solursh and Katow, 1982). Characterization of sulfate- thyme migration in over 93% of untreated control emlabeled glycosaminoglycans was also carried out as de- bryos. At lower concentrations (e.g.,
SOLURSH,MITCHELL, AND KATOW
Cell Migration
in Sea Urchin Embryos
327
a FIG. 1. Phase-contrast micrographs of the normal mesenchyme blastula (a) and an embryo treated with 2 mM @-D-xyloside, in which primary mesenchyme cell migration does not occur (b). The ingressed primary mesenchyme cells are migrating along the inner blastocoel wall toward the animal pole (A) in normal embryos (arrows in (a)) while those in the xyloside-treated embryo have ingressed but remain piled up at the vegetal plate (V) (arrows in (b)) X1100.
side. In the best case, only 55% of the embryos reared in sulfate-free seawater failed to undergo normal primary mesenchyme cell migration. In contrast to these treatments, culture of embryos in 2 or 4 mM nitrophenyl galactopyranoside, D-galactose, or D-xylose had no observable effect on development including primary mesenchyme cell migration. Morphology of the Blastocoel 2Matm.x Normal mesenchyme blastulae. By transmission electron microscopy, the blastocoel material appears as a complex of 15- to 30-nm granules and ca. 30-rim-diameter fibers (Fig. 2). The granules are often clustered, among which short, thin fibers about 4 nm in diameter are observed (Fig. 2, inset). The basal lamina also contains 15to 30-nm-diameter granules. These are localized on the blastocoel side of the basal lamina, while a fibrous material about 6 nm in diameter forms a network on the ectodermal side of the basal lamina (Fig. 2b). These fibers in the basal lamina appear morphologically different from the 4-nm fibers in the blastocoel, while the granules in the two locations are not morphologically distinguishable. The granules are also observed on the surfaces of cell processes extending from primary mesenchyme cells (Fig. 3), whether or not they are near the basal lamina. Xyloside-treated mesenchyme blastulae. Treatment with xyloside has a morphological effect on the blastocoel matrix. By transmission electron microscopy, the 30-nm granules are much less numerous than in normal embryos (Fig. 4) and clusters of the granules are seldom observed (Fig. 4a). On the other hand, the 4-nm-diameter fibers are often seen. The 30-nm granules are also less
numerous in the basal lamina, although the fibrous component is conspicuous (Fig. 4b). Effect of Xyloside on Sulfated Proteoglycans in the Blastocoel As considered, in many systems fi-xyloside stimulates the synthesis of some glycosaminogylcans (GAG). However, these are not protein bound but are free GAG chains. In contrast, treatment of S. purpuratus with 2 mM P-D-xyloside depresses the incorporation of 35SOzinto ethanol precipitable, nondialyzable material in the blastocoel matrix, where treated preparations ranged from 88 to 60% of control embryos expressed as CPM per embryo. The proportions of the various types of sulfated GAG present in Pronase digests were indistinguishable between normal controls and xyloside-treated embryos. As reported previously for normal embryos (Solursh and Katow, 1982), about 16% of the 35SOi- is found in chondroitin-6-S04, 37% in dermatan sulfate, 12% in heparan sulfate, 20% in glycopeptides, and the remainder still uncharacterized (Table I). Autofluorograms after sodium dodecyl sulfate-polyacrylamide gel electrophoresis of the sulfate-labeled blastocoel matrix from normal and xyloside-treated embryos on 7.5% gels are qualitatively indistinguishable from each other (not shown). Since proteoglycans are the major sulfated component in the blastocoel (Solursh and Katow, 1982) and these are likely targets of P-D-xyloside, the proteoglycans are examined in more detail. The size of the sulfate-labeled blastocoel components is analyzed by chromatography on a Sepharose CL-2B column using a dissociative solvent. As reported previously (Solursh and Katow, 1982),
328
DEVELOPMENTALBIOLOGY
VOLUME118, 1986
FIG. 2. TEM of normal mesenchyme blastula. The blastocoel material is composed of thin, long fibers (arrows in (a)) and granules, which are ca. 15-30 nm in diameter (large arrows in inset). These granules are associated with thin, short fibers (small arrows in inset). The basal lamina, on the other hand, is also composed of a fibrous material (thin arrows in (b)), which is morphologically different from the other fibrous structures in the blastocoel, and the 15- to 30-nm-diameter granules are attached to the blastocoel side of the basal lamina so that the basal lamina is polarized. Open arrows in (b) indicate the thin, short fibers of the blastocoel material. (a) X63,000. Inset, X145,000. (b) X128,000.
the majority of the label from normal embryos is eluted in the void volume (at least 60%, Fig. 5). Much of this label is sensitive to chondroitinase ABC pretreatment (Fig. 5), as reported earlier (Solursh and Katow, 1982). In contrast, only 10% of the % from the blastocoel matrix of xyloside-treated embryos was in the voided fractions. The majority of the label was included with about 44% having a Kay of 0.33. In this case too, the majority of the label is sensitive to chondroitinase ABC pretreatment, with a minor resistant component remaining in the voided fractions. These results suggest that while the overall GAG composition of blastocoel matrix is
similar in normal and treated embryos, its organization in proteoglycans is drastically altered. Based on results in other systems, one possibility is that xyloside treatment causes the formation of GAGdeficient proteoglycans and free GAG chains. The latter would be expected in the included fractions of the CL2B column. This possibility was examined by further analysis of the included fractions in Fig. 5 (fractions 65-85). While the majority of the counts are precipitated with cetylpyridinium chloride, only 10% of these counts were precipitated in cold TCA, suggesting that this material is largely in the form of free GAG chains. Based
SOLURSH, MITCHELL, AND KATOW
Cell Migration
in Sea Urchin
Embryos
329
on sensitivity to chondroitinase AC and ABC (Fig. 6), this material appears to be a mixture of chondroitin sulfate [chondroitin-6-SO1 based on earlier studies (Solursh and Katow, 1982)] and dermatan sulfate, in a ratio of 1:1.4, respectively. A sulfated fucogalactan-protein conjugate has been described in the extracellular matrix of sea urchin embryos (Akasaka and Terayama, 1982; 1983; 1934). Blastocoel material, from normal and xyloside-treated embryos, prepared after metabolic labeling with [3H]fucose was analyzed by chromatography on a Sephacose CL2B column (not shown). In this case, the elution patterns of samples from normal and xyloside-treated embryos were similar. About 30-45s of the recovered label was in the voided fractions and 45-58s was included. These results suggest that fucose rich components in the blastocoel are relatively resistant to xyloside treatment. DISCUSSION FIG. 3. TEM of primary mesenchyme cell process extended onto the basal lamina (BL) which is associated with 15- to 30-rim-diameter granules (arrows).
The morphological features of the blastocoel matrix in the mesenchyme blastula stage of S. purpuratus are
FIG. 4. TEM of mesenchyme blastula treated with 2 mM xyloside. The 15- to 30-nm-diameter granules are drastically reduced from both the blastocoel material (large arrows in (a)) and the basal lamina (large arrow in (b)) so that the short, thin fibers of the blastocoel material (small arrows in (a)) and the fibrous material of the basal lamina (small arrows in (b)) become more apparent than in the normal embryo (Fig. 2). These fibers appear intact morphologically. (a) X120,000 (b) X128,000.
330
DEVELOPMENTAL BIOLOGY
VOLUME 118, 1986
TABLE 1 S&-LABELED MATERIAL&JIN PRONASE DIGESTS OF BLASTOCOELMATRIX FROM NORMAL AND XYLOSIDE-TREATED MESENCHYME BLASTULAE Normal (%I Sulfated glycoprotein” Heparan sulfate* Chondroitin-6-SO,” Dermatan Sold Uncharacterized
20 12 16 37 15
Xyloside-treated (%) 23 12 16 36 13
Note. These results are typical of three separate trials. ‘The percentage of the total counts in the included fractions on a Sephadex G-50 column. bThe percentages of the total counts that are sensitive to nitrous acid treatment prior to chromatography on Sephadex G-50. “The percentages of the total counts that are sensitive to chondroitinase AC and cochromatograph with 2-acetamide-2-deoxy-3-0(fi-D-gluco-4-enepyranosyluronic acid)-6-sulfo-D-galactose standard. dThe percentages of the total counts that are sensitive to chondroitinase ABC and cochromatograph with either 2-acetamide-e-deoxy3-G-(fl-D-gluco-4-enepyranosyluronic acid)-D-galactose 4,gdisulfate or -4 or -6-sulfo-D-galactose minus the chondroitinase AC-sensitive counts.
generally like those described earlier in L pict~ (Katow and Solursh, 1979). The 15 to 30-nm granules are a prominent component and are present in the blastocoel on the blastocoel side of the basal lamina and on filopodia of migrating primary mesenchyme cells. They have the localization that one would expect for a component that is involved in the attachment of cells to the substratum. Notably, these granules are present at sites of cell attachment to the basal lamina (Katow and Solursh, 1981). Unlike S. pwpuratus, the thin fibrous component of L. pictus is not observed until the onset of gastrulation. Based on the morphology of these granules and their affinity for ruthenium red (Katow and Amemiya, 1986), they are likely to be proteoglycan granules (Hay, 1978). In addition to the association of the granules with sites of mesenchyme cell attachment, treatments that interfere with their accumulation inhibit cell migration. Sulfate deprivation of L pictus embryos permits primary mesenchyme cell ingression but prevents their migration (Karp and Solursh, 1974). The cells appear to be unable to form stable cell attachments to the basal lamina (Katow and Solursh, 1981). This treatment results in a paucity of the 15- to 30-nm granules on the basal lamina and in the blastocoel (Katow and Solursh, 1979). Sulfate deprivation of Clypeaetcrja~icus embryos inhibits the synthesis of dermatan sulfate (Yamaguchi and Kinoshita, 1985). Xyloside treatment, which is presumably a more specific inhibitor of the synthesis of normal proteoglycans than sulfate deprivation, has a similar effect. Primary mesenchyme cells ingress but fail to migrate (Kinoshita
"0
30
40
SO
60
70
n "
60
60
100
FIG. 5. S&-labeled hlastocoel matrix analyzed on a Sepharose CL2B column. S%-labeled blastocoel matrix from normal (0) or xylosidetreated (0) mesenchyme blastulae was analyzed on an analytical Sepharose CL-2B column with 4 Mguanidine HCl as solvent. At least 60% of the S% from normal embryos was in the void volume (V,). On the other hand, only 10% of the Ss6 from xyloside-treated embryos was voided. The remainder was spread throughout the included region with about 44% eluting with a K., of 0.33. If the samples were treated with chondroitinase ABC and passed through Sephadex G-50 prior to chromatography on Sepharose CL-LB, only a small percentage of the original counts from normal (0) and xyloside-treated (m) embryos is recovered. This material is largely restricted to the voided fractions. These results are typical of two trials.
ML
FIG. 6. Chondroitinase sensitivity of included fractions of blastocoel matrix from xyloside-treated embryos. S”-labeled blastocoel matrix from xyloside-treated embryos was analyzed on a dissociative CL-2B column and the major included fractions (fractions 65-85 in Fig. 5) were pooled, dialyzed, lyophilized, and run on a Sephadex G-75 column without prior treatment (0) or with prior treatment with either chondroitinase ABC (m) or chondroitinase AC (A). While 90% of the counts in the untreated sample are voided, 94% of the counts were in the Vt fractions after treatment with chondroitinase ABC. After treatment with chondroitinase AC, 38% of the counts are still voided with the remainder included or in the V,. Recovery was greater than 80% in all cases.
SOLURSH, MITCHELL, AND KATOW
Cell Migration
and Saiga, 1979; Akasaka et al, 1980; and Fig. 1). The present study shows that treatment results in a selective reduction of the 15 to 30-nm-diameter granules. Kinoshita and Saiga (1979) found that the inhibitory effects were observed only with P-xyloside and not with the (Yisomer, in contrast to the present study where both isomers are effective. This difference could be the result of differences in the ability to convert the cyisomer to the active /3 form. The specificity of xylosides is suggested by the absence of any detectible effect by D-galactose and fi-D-galactopyranoside. Xylose is also ineffective and is much less active in initiating the synthesis of glycosaminoglycans in other systems (Galligani et ak, 1975; Gibson and Segin, 1977). Other species differences include the relative resistance of primary mesenchyme cell migration in S. purpuratus to sulfate deprivation compared to L. pi&m and the relative resistance of primary mesenchyme cell migration in L. pictus to xylosides (unpublished observation). The biochemical comparisons of %S-labeled blastocoel matrix from untreated and xyloside-treated embryos are consistent with a role of a chondroitin sulfate/dermatan sulfate proteoglycan in primary mesenchyme cell migration. While the proportions of the various types of glycosaminoglycans in the blastocoel remain unaffected by xyloside treatment, with over half of the radiolabeled sulfated material being in the form of dermatan sulfate and chondroitin-6-sulfate, the total incorporation is reduced significantly. This later result agrees with the earlier report of Kinoshita and Saiga (1979). It should be recognized that only blastocoel material is analyzed here and some free glycosaminoglycan chains formed in the presence of xyloside could be present in other embryo fractions. The sulfated proteoglycans considered in this report are clearly heterogeneous and require further characterization. Even after xyloside treatment there remains additional chondroitinase ABC-sensitive label. Xyloside treatment appears to reduce the size of the major sulfated proteoglycans in the blastocoel matrix. Over 60% of the label from untreated embryos is voided from a Sepharose CL-ZB column and appears to be larger than 40 X lo6 Da. Much of the label is sensitive to chondroitinase ABC, as reported earlier (Solursh and Katow, 1982). Only 10% of the label is voided when material from xyloside treated embryos is examined and a large proportion of the 35SOf- is included with a K,, of 0.33. Most of this later material is sensitive to chondroitinase ABC and appears to consist of free glycosaminoglycan chains, as one would expect after xyloside treatment. The relative increased proportion of chondroitin sulfate to dermatan sulfate (1:1.4) in the included fractions compared to that in the whole extract (1:2.3) suggests heterogeneity in the dermatan sulfate containing proteoglycans. While dermatan sulfate is normally xyloside
in Sea Urchin Embryos
331
linked (Akiyama and Seno, 1981), this might not be the case for all sea urchin dermatan sulfate (Oguri and Yamagata, 1978). The present results are consistent with the existence of xyloside linked dermatan sulfate and chondroitin sulfate-rich proteoglycans as well as dermatan sulfate proteoglycans that do not contain an Oglycosidic bond. While this possibility remains to be investigated, xyloside linked chondroitin sulfatejdermatan sulfate-rich proteoglycans may be specifically involved in cell migration. It is not yet known if a single proteoglycan containing both chondroitin sulfate and dermatan sulfate mediates mesenchyme cell attachment during migration. This work was supported
by NIH Grant HD16549. REFERENCES
AKASAKA, K., AMENIYA, S., and TERAYAMA, H. (1980). Scanning electron microscopical study of the inside of sea urchin embryos (Pseudo-
eentrotus depress+&. Exp. Cell Res. 129,1-13. K., and TERAYAMA, H. (1982). Sulfated fucogalactan-protein conjugate present in the EDTA-extract from Anthocidaris embryos (mid-gastrula). J. Fat. Sci Univ. Tokgo. 15,181-189. AKASAKA, K., and TERAYAMA, H. (1983). Sulfated glycan present in the EDTA extract of Hemicentrotus embryo (mid-gastrula). Exp. Cell. Res, 146,177-185. AKASAKA, K., and TERAYAMA, H. (1984). A proteoglycan fraction isolated from the EDTA extract of sea urchin (Hemicentrotus pulcherrimus) gastrulae stimulates reaggeration of dissociated embryonic cells. Exp. Cell Res. 150, 226-233. AKIYAMA, F., and SENO, N. (1981). Linkage regions between dermatan polysulfates and peptides. B&him. Biophys. Acta 674,280-296. ANDERSON, T. (1951). Technique for preservation of 3-dimensional structure in preparing specimens for the E. M. Trans. N. Y. Acad sci 13,130-134. DAWSON, R. M. C., ELLIOTT, W. H., and JONES, K. M. (1969). “Data for Biochemical Research.” p. 508. Oxford Univ. Press, London/New York. FUKUNAGA, Y., SOBUE, M., SUZUKA, N., KUSHIDA, H., and SUZUKI, S. (1975). Synthesis of a fluorogenic mucopolysaccharide by ehondrocytes in cell culture with I-methylumbelliferyl @D-xyloside. B&him. Biophys. Acta 381,443-447. GALLIGANI, L., HOPWOOD,J., SCHWARTZ, N. B., and DORFMAN, A. (1975). Stimulation of synthesis of free chondroitin sulfate chains by O-Dxyloside in cultured cells. J. Biol. Chem. 250,5400-5406. GIBSON, K. D., and SEGEN, B. J. (1977). The mode of action of 4-methylumbelliferyl @-D-xyloside on the synthesis of chondroitin sulphate in embryonic-chicken sternum. Biochem J. 168,65-79. HARKEY, M. A., and WHITELEY, A. H. (1980). Isolation, culture, and differentiation of echinoid primary mesenchyme cells. Rwx’s Arch. 189.111-122. HART, G. W., and LENNARZ, W. J. (1978). Effects of tunicamycin on the biosynthesis of glycosaminoglycans by embryonic chick cornea. J. AKASAKA,
Biol. Chem. 253,5’795-5801. E. D. (1978). Fine structure of embryonic matrices and their relation to the cell surface in ruthenium red-fixed tissues. Growth 42,399-423. KARP, G. C., and SOLURSH, M. (1974). Acid mycopolysaccharide metabolism, the cell surface, and primary mesenchyme cell activity in the sea urchin embryo. Deu. Biol. 41.110-123. KATOW, H., and AMEMIYA, S. (1986). Behavior of primary mesenchyme cells in situ associated with ultrastructural alterations of the blasHAY,
332
DEVELOPMENTALBIOLOGY
tocoelic material in the sea urchin, Anthocidaris crastipi~~~. Dev. Growth D@Y. 28.31-42. KATOW,H., and SOLURSH,M. (1979). Ultrastructure of blastocoel material in blastulae and gastrulae of the sea urchin Lytechinus pictus. J. Exp. ZooL 210,561-567. KATOW, H., and SOLURSH,M. (1980). Ultrastructure of primary mesenchyme ingression in the sea urchin Lytechinus pi&us. J. Exp. Zool. 213,231-246. KATOW, H., and SOLURSH,M. (1981). Ultrastructural and time-lapse studies of primary mesenchyme cell behavior in normal and sulfatedeprived sea urchin embryos. Exp. Cell Res. 136,233-245. KINOSHITA, S., and SAIGA, H. (1979). The role of proteoglycan in the development of sea urchins. Exp. Cell Res. 123,229-236. LAEMMLI, U. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 227,680685.
VOLUME118, 1986
OGURI,K., and YAMAGATA,T. (1978). Appearance of a proteoglycan in developing sea urchin embryos. Biochim Biophys. Actu 541.385-393. REYNOLDS, E. S. (1963).The use of lead citrate at high pH as an electronopaque stain in electron microscopy. J. CeUBiol. 17,208-212. SAITO, H., YAMAGATA,T., and SUZUKI, S. (1968). Enzymatic methods for the determination of small quantities of isomeric chondroitin sulfate. J. BioL Chem 243,1536-X42. SOLURSH,M. (1986).Migration of sea urchin primary mesenchyme cells. In “Developmental Biology: A Comprehensive Synthesis.” Vol. 2, pp. 391-431. Plenum, New York. SOLURSH,M., and KATOW,H. (1982). Initial characterization of sulfated macromolecules in the blastocoels of mesenchyme blastulae of Stmng&xentrotus purpuratus and Lytechinus pi&us. De-v.Biol 94, 326-336. YAMAGUCHI,M., and KINOSHITA, S. (1985). Polysaccharides sulfated at the time of gastrulation in embryos of the sea urchin Clypeaster japonicw. Exp. Cd Res. 159,353-365.