aggregation is the control of aggregate size. If reaggregation is not stopped by dilution when a certain clump size has been reached, monsters are produced.
Proc. Natl. Acad. Sci. USA Vol. 76, No. 1, pp. 288-292 January 1979 Cell Biology
Reconstitution of membranes and embryonic development in dissociated blastula cells of the sea urchin by reinsertion of aggregation-promoting membrane proteins extracted with butanol (cell surface recognition proteins/cell-cell binding/change of phenotype)
HANS NOLL*t, VALERIA MATRANGAt, DOMENICO CASCINOt, AND LETIZIA VITTORELLIt *Department of Biochemistry and Molecular Biology, Northwestern University, Evanston, Illinois 60201; and tfstituto di Anatomia Comparata, Via Archirafi, 20,
90123 Palermo, Italy
Communicated by Niels Kaj Jerne, September 11, 1978
ABSTRACT Blastula embryos of the sea urchin Paracentrotus lividus, when dissociated into single cells by exposure to Ca2+_ and Mg2+-free sea water, reassociate spontaneously to form aggregates capable of development to the final larval form (pluteus). This aggregation is prevented by Fab fragments obtained by immunization with purified membranes from blastula embryos. The inhibition was reversed by soluble proteins extracted with butanol from purified membranes or from intact cells. These extracts also strongly stimulated the rate of reaggregation of dissociated cells in the absence of Fab fragments. Exposure of dissociated cells to 2.5% (vol/vol) butanol removed completely the protein(s) responsible for reaggregation of the cells without impairing their viability. Reaggregation and embryonic development were completely restored to the extracted cells by readdition of the proteins extracted from either membranes or cells. Extracted cells from Paracentrotus could be reconstituted with proteins from Arbacia. The orderly integration of cells into tissues and organs during embryogenesis and the maintenance of this order in the adult as well as its breakdown in cancer are important phenomena that are poorly understood. Recent findings suggest that tissue-specific cell adhesion in development is mediated by specific recognition proteins in the membrane (1-7). Differentiation could thus be understood as the sequential expression of recognition proteins that, like antibodies, might be members of a general class (8). Cancer, on the other hand, might result from mutational alterations in these recognition proteins that lead to a loss of tissue-specific cell adhesion (invasiveness) in the mutated cell. The isolation and characterization of these recognition or contact proteins have been hampered by solubilization and detection problems, and in many cases the presence of these antigens on the cell surface has been inferred only from immunological methods (9, 10). Proteolytic degradation and denaturation by detergents have been encountered in attempts to isolate these glycoproteins from either culture fluids or purified membranes. Another problem is detection by a sensitive biological test. Assays based on promotion of cell aggregation require uptake of the active fraction by cells whose membranes have been partially depleted. On the other hand, immunological assays based on blocking cell contact by monovalent Fab fragments derived from antibodies raised against whole membranes are highly specific and do not require interaction of the active fraction with the cell (10, 11). However, because this test measures the extent to which the blocking action is reversed after absorption of the Fab fragment preparation with the fraction to be tested, the method is applicable only if the solubilized contact sites are preserved in a conformation that can be recognized by the antibody.
Many of these difficulties would disappear if it were possible to remove completely proteins associated with the function under study frort the membranes of living cells without impairing their viability. The active component could then be identified by testing for restoration of the deficient function after the reincorporation of the active component into the membrane. We have accomplished this with cells dissociated from the blastula embryos of the sea urchin Paracentrotus iividus. Sea urchins are particularly well suited for the study of specific aggregation in embryogenesis because, as established by the pioneering studies of Giudice and coworkers (12-16), dissociated cells readily reassociate to form morula-like structures able to complete development to the final pluteus stage. In this study we show that it is possible to treat dissociated blastula cells from P. lividus with butanol in such a way that they completely lose their ability to reaggregate. The ability for reaggregation was restored by the addition of a soluble protein fraction that had been extracted with butanol from purified membranes or from whole cells of normal blastula embryos. Apparently, the deficient membranes are fully reconstituted, because the aggregates formed upon treatment of the deficient cells with the membrane extract were able to differentiate and to complete their development into pluteuslike larvae. MATERIALS AND METHODS Dissociation of Embryos into Single Cells and Extraction with Butanol. P. lividus embryos were cultured and dissociated at the swimming blastula stage as described by Giudice and Mutolo (16). The cells were sedimented and resuspended in 50 vol of sea water for use in microtiter assays. For extraction, the packed cells were suspended in 20 vol of sea water containing 2.5% (vol/vol) 1-butanol and immediately centrifuged at 100 X g for 2 min. The supernatant was concentrated and dialyzed extensively against sea water. The cells were resuspended in 50 vol of sea water. All procedures were carried out at 15-18'C unless specified otherwise. Reaggregation Assay. Aggregation-promoting activity was determined in circular wells (10 mm wide, 5 mm deep) of microtiter plates. The total volume was always 0.1 ml and the cells (ca. 4 X 104) were added last in a volume of 0.01 ml. When embryonic development was to be followed, the samples were diluted after 1-2 hr in order to prevent furthef growth of the aggregates. Dilution was with 0.3 ml of sterile sea water containing antibiotics. The extent of aggregation was quantitated by visual inspection and photographic recording. Purification of Membranes and Extraction with Butanol. The detailed procedures will be published elsewhere. Briefly, the embryos were homogenized and the low-speed pellet (25,000 X g for 10 min) was dissolved in 10 mM Tris-HCl, pH
The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U. S. C. §1734 solely to indicate this fact.
tTo whom reprint requests should be addressed. 288
Cell Biology: Noll et al.
Proc. Natl. Acad. Sci. USA 76 (1979)
7.6, and centrifuged through two layers of sucrose (5 ml of 60%, 9 ml of 48%, wt/vol) in a Beckman SW-25 rotor at 20,000 rpm, 15 hr, 5YC. The material from the two interfaces was combined, diluted with H20, and sedimented at 40,000 rpm for 60 min, and the pellet was further purified by partitioning in the twophase system of Brunette and Till (17), using 10 mM Tris-HCl, pH 7.6, as buffer and no ZnC12. The pellicles at the interface were suspended in H20 and, after centrifugation as above, stored as pellets at -60°C. Electron microscopy showed uniform smooth vesicles. For extraction, the membrane pellets were suspended in 0.2 M Tris-HCl, pH 8, and mixed with 0.5 vol of 1-butanol at 4VC (18). The lower aqueous phase was collected after separation by centrifugation in a swinging-
bucket rotor, concentrated, and dialyzed against sea water. The yield from purified membranes corresponding to 14 ml of packed cells was 4.3 mg of protein. Aggregation-promoting activity was detectable at 3 jig/ml. Preparation of Fab Fragments. Two rabbits were immunized with a suspension of purified membranes from late blastula embryos (1 mg of protein per ml of saline). First injection: 2 ml intravenously; second, after 30 days: 0.5 ml intravenous, 1 ml in complete Freund's adjuvant at multiple intradermal sites. Two weeks later, 50 ml of blood was collected from each rabbit. Fab fragments were prepared from the pooled serum by either pepsin or papain digestion, essentially according to Nisonoff (19). The products were repurified by gel filtration on Sephadex G-150 and shown to have the expected s-values in a sucrose gradient. RESULTS Prevention of Reaggregation by Membrane-Specific Fab Fragments. Upon fertilization, sea urchin eggs within 48 hr develop in perfect synchrony through a series of morphologi-
cally distinct stages to the final larval form, called a pluteus (Fig. 1). During this period, in which complicated structures consisting of skeleton, digestive tract, and organs of locomotion are formed, total cell mass does not increase appreciably although cell number increases from one to several thousand. Giudice and coworkers have shown (12-16) that single cells derived from embryos of the earlier stages (blastula, gastrula) reaggregate spontaneously to form clumps that differentiate in a process that imitates normal embryonic development. However, the embryos resulting from reaggregation are usually less perfect and are morphologically more variable than the products of normal embryogenesis. Particularly critical in reaggregation is the control of aggregate size. If reaggregation is not stopped by dilution when a certain clump size has been reached, monsters are produced. If, on the other hand, the clumps are below a critical size, they are unable to complete development. The initial stages of reaggregation are detectable after a few minutes under the microscope by the formation of doublets, followed by larger chains in which contact between adjacent cells is visible by the formation of flattened boundaries. After
289
about 1-2 hr, clumps consisting of many cells appear (Fig. 2 b and i) that eventually detach from the substratum and swim around vigorously (Fig. 2d) with the aid of cilia on the outer cells that are beginning to form an epithelial layer (Fig. 2c) around the blastula-like structures (21, 22). Reaggregation of dissociated blastula cells was prevented by monovalent antibodies from rabbits immunized with purified membranes prepared from blastula embryos (Fig. 2 k and 1). Monovalent Fab fragments derived from purified IgG either by papain or by pepsin followed by reduction were equally active and completely prevented reaggregation at concentrations of 20-40 ,gg/0. 1-ml sample. Because of the sensitivity of the cells to changes in their normal sea water medium, we found it necessary to dialyze all materials against sea water prior to testing. Inhibition of reaggregation required monovalent fragments. By contrast, divalent fragments F(ab')2 resulting from digestion with pepsin caused the cells to agglutinate (not shown). However, the agglutinated cells, in contrast to the physiological aggregates, maintained their rounded shape until they eventually began to disintegrate. Addition of pepsin to the dissociated cells did not prevent reaggregation. Dissociation of Blastula Embryos by Fab Fragments. Addition of Fab fragments to blastula and gastrula embryos caused dissociation into single cells within several hours. The embryos rapidly became immobile and gradually lost their characteristic shape as they were converted into clumps resembling piles of single cells with rounded contours instead of the cobblestone pattern of cells in contact with each other. The amount of debris and single cells increased with time and antibody concentration. After a few hours the embryos were converted into mostly single cells that remained intact for over 12 hr. By contrast, divalent F(ab')2 fragments did not promote dissociation even though they caused a similar initial loss of ordered structure, which was followed by cell lysis (10). Reversal of Antibody-Induced Inhibition of Reaggregation by Membrane Extract. We next tested whether components extracted from purified membranes with butanol and dialyzed against sea water could abolish the inhibitory effect of monovalent Fab fragments. A titration was carried out in which butanol extract was added in increasing concentrations to two different sets of wells in microtiter plates, one containing 62 Aig and the other 124 ,ug of Fab fragments. Dissociated cells were then added to each series. The results in Table 1 show that the butanol extract is indeed capable of completely reversing the aggregation-inhibiting effect of the Fab fragments and that the concentration of butanol extract required for neutralization increases with the concentration of antibody. Moreover, the reaggregated cells were able to develop into embryos indistinguishable from those produced by spontaneous reaggregation in the absence of antibodies. Promotion of Reaggregation by Direct Interaction of Butanol Extract with Dissociated Cells. Subsequent tests showed that the activity of the butanol extracts could be assayed Table 1. Reversal of Fab-induced inhibition of reaggregation by butanol extract from membranes
Fab,,ug 0 b
a
c
d
0 62 124
0 0 0
Butanol extract, ,ug of protein per 0.1-ml culture 6 12 21 3 0.15 0.3 1.5 0.03 1+ ND ND
2+ 0 0
3+ 0 0
3+ 0 0
4+ 1+ 0
4+ 2+ 1+
ND ND 3+ 4+ 2+ 2+
Extent of reaggregation is estimated on a scale from 0 (no aggre9
f
e
h
development (20): freeswimming blastula (a, b), gastrula (c, d), prism (e, f), pluteus (g, h). FIG. 1.
Stages
of
early
sea urchin
gation) to 4+ (maximal aggregation) observed after 1 hr. Fab was prepared by the papain method. Butanol-extracted cells were used in the experiments with no Fab; normal cells were used for all experiments with Fab. ND, not done.
Cell Biology: Noll et al.
290 a
?4;a~~~t
;
Proc. Natl. Acad. Sci. USA 76 (1979)
:
