The effects of disrupting cell interactions in early development were investigated by examining the accumulation of a primary mesenchyme specific transcript (SM50) and an aboral ectoderm-specific transcript (Spec 1) in cultures of sea urchin embryos that were dissociated at early stages and then cultured in CFSW. The expression of both SM50 and Spec 1 is temporally correct and remains restricted to the appropriate cell types, even if the embryo is dissociated as early as the 2-cell stage and maintained as a suspension of single cells. This result is consistent with the idea that the specificity of expression of these two genes, each characteristic of different lineages, is strongly regulated by information in the egg. Average SM50 expression is half that of intact embryos, but Spec 1 expression is very low, only 10–20 % of intact controls, suggesting some differences in the response of the two genes to lack of close cell interactions.

Cell interactions are known to be important in developmental processes and are probably used as a developmental strategy to varying extents in all animals. Cell associations play crucial roles in the development of both protostomes and deuterostomes (Davidson, 1986; Gurdon et al. 1984; Gurdon, 1987), although the nature and mechanisms of such interactions are unknown and deserve further study.

The overall pattern of determined tissues in sea urchin embryos is established early in development through a set of interactive processes. For example, isolated blastomeres from 2- or 4-cell-stage sea urchin embryos are capable of forming complete 1/2 or 1/4 size larvae (Driesch, 1891). Furthermore, transplantation experiments have suggested that some cells can induce others to take on new fates (reviewed by Horstadius, 1973; Davidson, 1986, 1989; Wilt, 1987). Such results argue for the proposition that some cells of the early sea urchin embryo receive and react to cues from neighboring cells. However, previous experimental manipulations in sea urchin embryos have been difficult to interpret (Wilt, 1987) due to the reliance on solely morphological criteria in assessing the resulting state of cellular differentiation. For example, if the appearance of a particular morphological character requires the mechanical processes of gastrulation, then manipulations inhibiting gastrulation would preclude the assessment of differentiation in some tissues. Fortunately, tissue-specific markers are being developed for use in sea urchin embryos which allow unambiguous identification of differentiated cell types (Davidson, 1986; Angerer and Davidson, 1986; McClay and Wessel, 1985; Wilt, 1987).

The experiments reported here examine the role of cell interactions in sea urchin development by dissociating embryos at early stages, preventing subsequent reassociation, and then looking for the initial appearance and accumulation of tissue-specific transcripts. The two tissue-specific markers used in these experiments are SM50 and Spec 1. SM50 is a 50 kD spicule matrix protein that is the principal matrix glycoprotein of the endoskeletal spicule of the sea urchin larva. It is synthesized from a 2.2 kb RNA present only in primary mesenchyme cells, a group of approximately 32 cells which give rise to the larval skeleton (Benson et al. 1987; SucovetaZ. 1987). Primary mesenchyme cells are descendants of the micromeres, a set of four small cells formed at the vegetal pole upon the 4th cleavage division. SM50 expression is therefore both lineage and tissue specific.

Spec 1 codes for a 16 kD troponin C-like protein that contains Ca2+-binding domains. Spec 1 transcripts (1.5 kb) are initially present at low levels in all cells of the early embryo but accumulate to high levels in about 200 aboral ectoderm cells of the mesenchyme blastula (500-cell-stage) embryo (Carpenter et al. 1984; Cameron et al. 1987). Spec 1 is also tissue specific since it is eventually expressed only in aboral ectoderm; however, several different blastomeres of the 16-cell-stage embryo, including some from both the animal and vegetal hemispheres, contribute to aboral ectoderm, and therefore it is not a lineage-specific marker (Cameron et al. 1987).

We have found that embryos dissociated as early as the 2-cell stage accumulate SM50 transcripts at the appropriate time (compared to intact control cultures) and in the appropriate cell type. Spec 1 expression is very low, but it also begins to accumulate at the appropriate time and in the appropriate cells.

Dissociation at 4-cell stage

Gametes of Strongylocentrotus purpuratus were obtained, eggs fertilized, and embryos cultured by conventional methods (Hinegardner, 1967; Schroeder, 1986). Fertilization was carried out in 5 mM-sodium p-aminobenzoic acid in sea water (Hall, 1978) to soften fertilization membranes, which were removed 30 min after fertilization by passage through a 55 pm Nitex screen. Embryos were then allowed to settle 2 times in Millipore-filtered (0.45 μm) sea water (MPFSW), 2 times in calcium-/magnesium-free sea water (CMFSW), and then stirred (60 revs min-1) in calcium-free sea water (CFSW): CMFSW, 3:1, until the 4-cell stage, by which time the blastomeres were usually dissociated. In the occasional experiment where blastomeres still remained attached, the embryos were settled 1-2 more times in CMFSW and swirled vigorously in a 500 ml beaker. Once separated, blastomeres were stirred at 60 revs min-1 and cultured in CFSW or CFSW: CMFSW 3:1 at 0.2·1 ×104 initial cells ml-1 dilution. Every 2h thereafter for the next 4-6 h, cultures were poured through 35 μm Nitex to prevent reaggregation. Mesomeres and macromeres were separated from micromeres on sucrose gradients essentially according to the methods of Kitajima and Matsuda (1982) and then cultured by stirring in CFSW (without horse serum). Culture densities were determined by haemacytometer cell counts.

RNA isolation

RNA was isolated based on a method devised by March (1985) and modified by Bruce Brandhorst (personal communication). Cells or embryos were pelleted and washed 2 times with CMFSW, resuspended in 5 packed cell volumes of guanidine isothiocyanate (GuSCN) homogenization buffer (4.5M-GUSCN, 50 mM-sodium citrate pH 7, 5% beta-mercaptoethanol (βME), 0.5% Sarcosyl) and homogenized vigorously 15–20 times in a tight-fitting homogenizer. After addition of 0.6 volumes of 100 % ethanol, homogenates were stored overnight at —20°C. The precipitate was collected at 10000 g for 10 min and the pellet dissolved in 8M-guanidine hydrochloride (GuCl), 50 mM-sodium citrate pH 6, using the same volume as the homogenization buffer. Occasionally it was necessary to draw the pellet through a syringe needle in order to facilitate dissolution of the pellet. The material dissolved in GuCl was extracted twice with phenol: chloro-form:isoamyl alcohol (50:48:2), once with chloroform:isoamyl alcohol (96:4), and then precipitated with 0.6 volumes ethanol overnight at −20°C. The precipitate was collected at 10000g for 10 min and the pellet resuspended in 0.3M-sodium acetate, pH5.2 and again precipitated with 2.5 volumes ethanol at −20°C overnight. The pellet was collected the same as above, dissolved in H2O and the amount of RNA determined spectrophotometrically.

Clones and probes

The SM50 probe is derived from a 1.3 kb sequence that was isolated and characterized by Sucov et al. (1987) and Benson et al. (1987) and subcloned in both orientations into the pSP65 transcription vector (Melton et al. 1984). The Spec 1 probe is derived from a (EcoRI to Ba/trHI site) sequence isolated and characterized by Carpenter et al. (1984) from which a 260 bp sequence was isolated and subcloned into the pGEM 1 and 2 transcription vectors (Promega Biotec, Madison, WI). The subcloned sequence is within the open reading frame of Spec 1. The probe for Cy I actin consists of a 780 bp RI to WmdIII fragment from the 3′ untranslated portion (Lee et al. 1984) which was subcloned into the pGEM2 transcription vector by Andrew Cameron. A 512 nt region of S. purpuratus mitochondrial 16S ribosomal RNA (Wells et al. 1982; Eldon et al. 1987) was used for quantification of blots. SP6 or T7 polymerase (Promega Biotec and Bethesda Reasearch Laboratories) was used to synthesize all RNA probes essentially according to Melton (1984).

RNA blots

5 μg RNA were loaded per lane on agarose gels following the method of Maniatis (1982) and transferred to nitrocellulose (Thomas, 1980). RNA blots were hybridized exactly according to Benson et al. (1987) using 32P-UTP labeled riboprobes synthesized from the pSP65 transcription vector for SM50 and the pGEM transcription vector for Spec 1 and Cyl. Gel blots were tested for probe specificity by washing with 0.05 μg RNase ml-1 in 2×SSC (0.3M-sodium chloride, 0.03M-sodium citrate) for 15 min at 37°C. Blots were exposed to flashed X-Ray film (to obtain a linear response) at −80°C and densitometry was performed on the resulting autoradiograms. Equal loading of RNA per lane was confirmed by one of two methods: acridine orange staining of gels prior to transfer or by hybridization of blots with a mitochondrial 16S rRNA probe.

In situ hybridization

Cells were prepared for in situ hybridization either by following standard fixation and histological sectioning procedures (Cox et al. 1984) or by spreading whole cells on polylysine-coated slides followed by fixation and dehydration. The protease treatment, hybridization steps, RNase A digestion and autoradiography were all performed according to Cox et al. (1984). 3H-labeled single-stranded RNA probes were synthesized from the pSP65 transcription vector for the SM50 probes (in both orientations) and pGEM for the Spec 1 probes (in both orientations) using 3H-UTP; the probes had a specific activity of 10 Ci mmole-1 of probe nucleotide. NTB emulsion-coated slides were exposed for 18 or 28 days at 4 °C before development. 35S-labeled single-stranded RNA probes were also synthesized from the Spec template following the method of Angerer et al. (1987). Hybridizations and washes were carried out according to that protocol with the following exceptions: slides were prehybridized for 4h at 47°C in hybridization buffer without probe; hybridization and prehybridization buffers contained 100pM-UTP; all post-hybridization washes contained 10mM-βME. The 35S slides were exposed for 11–13 days at 4°C. Spec 1 and SM50 sense probe hybridizations (both 3H and 35S) showed only background levels of grain counts over cells. Cells hybridized to antisense probes were scored positive if grain counts were at least twofold greater than background levels determined for each experiment and probe.

Dissociated early embryos form viable cell cultures

We examined the role of blastomere interactions in the regulation of tissue-specific gene expression by radically altering normal cellular interactions. A simple and drastic method is to prevent any cell interaction whatsoever by dissociating the embryos at various early stages and then keeping subsequent daughter cells from reassociating. Embryos were dissociated by decreasing Ca2+ and Mg2+ levels in the sea water, combined with modest shear. Cells continue to divide (if Mg2+ is present in the sea water), presumably using internal Ca2+ stores for their usual cellular activities.

To verify that the dissociated cells were actually single cells and remained so, the behaviors of cells in dissociated cultures were carefully monitored. In the majority of the experiments reported here, embryos of S. purpuratus were dissociated at the early 4-cell stage (Fig. 1A) before the second division was completed, and thus, the blastomeres initially came apart in pairs (Fig. IB). When cultured in CFSW and stirred at 60revsmin-1, these cell pairs fell apart prior to the next division cycle. During the 6–8 h following dissociation, nascent daughter pairs briefly remained attached until the next cleavage cycle (approximately one hour later) at which time they dissociated from one another, thus forming a true single cell suspension. All cells appeared identical in their rounded morphology and lack of cilia. No overt signs of differentiation, such as pigmentation or spiculogenesis, were observed. Occasionally cells clumped, presumably due to continued hyalin synthesis during this period. However, simple swirling or passing of the culture through 35 /rm Nitex restored the culture to a single-cell suspension and this was performed hourly or as needed. After the initial 8h, mechanical disruption was no longer necessary to maintain the dissociated state. After 24h of culture in CFSW (Fig. 1C), well over 90% of the cells in culture consisted of normally sized, single cells. Usually about 2 % of these cells (the upper limit was 5 %) were very large and were presumably uncleaved blastomeres (detectable by their relatively large size) and less than 5 % of the cells existed in small clumps of 5–10 cells. The large cells and clumps were eliminated by allowing them to settle out of the culture for 15–30 min prior to harvesting of dissociated cultures for RNA; thus, their effects on our results are negligible.

Fig. 1.

Micrographs of embryos and dissociated cell cultures. 4-cell-stage embryos (A), dissociated blastomeres immediately following dissociation at the 4-cell stage (B) and the same dissociated cells after stirring for 24 h in CFSW (C). Scale bar represents 150 μm.

Fig. 1.

Micrographs of embryos and dissociated cell cultures. 4-cell-stage embryos (A), dissociated blastomeres immediately following dissociation at the 4-cell stage (B) and the same dissociated cells after stirring for 24 h in CFSW (C). Scale bar represents 150 μm.

Direct evidence for division of dissociated cells in CFSW was obtained by counting the cells during culture. Fig. 2 shows the logarithmic increase in cell number in control (solid lines) and dissociated (dashed lines) cultures. The latter accumulate nearly the normal number of cells, though by 24h they have, on the average, undergone one less cell division than the control cultures. Greater than 95% of the cells in dissociated cultures, and in control embryos dissociated at the mesenchyme blastula stage, exclude trypan blue. Furthermore, dissociated cells and intact embryos show similar numbers of labeled cells (averaging about 90 %) after a short pulse of [3H]leucine followed by autoradiography.

Fig. 2.

Comparison of cell numbers in three intact control cultures (solid lines and symbols) and three dissociated cell cultures (dashed lines, open symbols) from 0 to 24 h after fertilization. Initial cell densities were normalized to facilitate comparisons.

Fig. 2.

Comparison of cell numbers in three intact control cultures (solid lines and symbols) and three dissociated cell cultures (dashed lines, open symbols) from 0 to 24 h after fertilization. Initial cell densities were normalized to facilitate comparisons.

It has been known that exposure to CFSW for several divisions does not cause lack of viability (Horstadius, 1973). Furthermore, previous disaggregation-reaggregation experiments showed little effect on cell viability, although most of those experiments consisted of disaggregation followed soon thereafter by reaggregation (reviewed by Giudice, 1986). We carried out experiments in which Ca2+ was restored to the medium 5 h after dissociation, equivalent to the 64-to 128-cell-stage in controls, and cells were allowed to reaggregate without stirring for 72 h. Abnormal but recognizable plutei were formed, complete with ectoderm, gut and spicules. Hurley et al. (1989) report that cells cultured in CFSW for 24 h and then reaggregated had normal levels of [35S]methionine incorporation, although such reaggregates do not undergo any morphological signs of development.

Finally, dissociated cells accumulate mRNA. We followed the accumulation of ubiquitous cytoskeletal actin transcripts during development of intact and control dissociated cultures. During normal development Cy I actin transcripts are present at very low levels in the 2-cell-stage embryo, then accumulate in virtually all cell types during the second half of cleavage due to transcriptional activation between 9 and 12 h. At 24 h (mesenchyme blastula stage), the mRNA begins to disappear in intact embryos from aboral ectoderm, and to a lesser extent from oral ectoderm, while continuing to accumulate in presumptive secondary mesenchyme and gut (Cox et al. 1986). Hence, Cy I is a good indicator of transcriptional activation throughout all cells of the cleavage stage embryo up until the mesenchyme blastula stage, the time when our experiments terminate. Embryos dissociated at the 4-cell stage and cultured in CFSW until 24 h postfertilization accumulate near normal levels of Cy I mRNA (116±23 % in 3 experiments) compared to control mesenchyme blastula levels. Hurley et al. (1989) have obtained similar results.

Embryos dissociated during early cleavage stages express tissue-specific genes

We wished to determine if new gene expression characteristic of differentiated tissues would occur in the absence of normal cell contact. Embryos were dissociated at earlier and earlier stages. Embryos dissociated at the 500-cell stage (mesenchyme blastula) and cultured in CFSW to the equivalent of the gastrula stage (48 h) expressed both SM50 and Spec 1 RNAs (Fig. 3, lane d), as did embryos dissociated at the 128-cell, 16-cell, or the 2-cell stages (Fig. 3A, lanes c,b,a). Densitométrie measurements performed on the autoradiograms shown in Fig. 3 revealed that SM50 expression in all four dissociated cultures, regardless of the stage of dissociation, attained over 50% of the transcript level found in the control mesenchyme blastula embryo (Fig. 3A, lane e). In contrast, the level of Spec 1 expression in dissociated cultures of mesenchyme blastulae was less than 35 % of that found in control (undissociated) mesenchyme blastulae, and was increasingly diminished in embryos dissociated at earlier stages (Fig. 3B). Dissociation at the 128-, 16- and 2-cell stages yielded 22 %, 13 %, and 9 %, respectively, of the control mesenchyme blastula Spec 1 transcript level (Fig. 3B lanes c,b,a).

Fig. 3.

RNA blots of SM50 (A) or Spec 1 RNA (B) present in cultures from embryos dissociated at different times after fertilization. Embryos dissociated at the 4-cell (a), 16-cell (b), 128-cell (c) and mesenchyme blastula (d) stages and then cultured in CFSW until intact, control embryos reached the gastrula stage. Lane (e) contains control mesenchyme blastula RNA.

Fig. 3.

RNA blots of SM50 (A) or Spec 1 RNA (B) present in cultures from embryos dissociated at different times after fertilization. Embryos dissociated at the 4-cell (a), 16-cell (b), 128-cell (c) and mesenchyme blastula (d) stages and then cultured in CFSW until intact, control embryos reached the gastrula stage. Lane (e) contains control mesenchyme blastula RNA.

Expression of SM50 and Spec 1 in dissociated cultures occurs at the normal time

We examined the kinetics of accumulation for both SM50 and Spec 1 transcripts in dissociated cultures (Fig. 4). Because of the variability in time and amount of expression of these transcripts found in different egg batches, comparisons were made to control embryos from the same batch of eggs. Exceedingly low levels of SM50 transcription are detectable at the early blastula stage and become more apparent at the mesenchyme blastula stage when both control embryos and dissociated cells produce markedly increased levels of SM50 transcript (Fig. 4A, lanes d,d′)-This pattern of accumulation is characteristic of SM50 (Killian and Wilt, 1989). Unlike SM50, low levels of maternally derived Spec 1 transcript are present at the 4-cell and 7h midcleavage stage (Fig. 4B, lanes a to b′). At the 12 h early blastula stage a zygotic accumulation of Spec 1 transcripts is evident (Fig. 4B, c-c′) though continued later accumulation in dissociated cultures is very low. Densitométrie evaluation of autoradiograms from similar experiments (Fig. 5) shows that for both SM50 and Spec 1, transcript accumulation in the dissociated cultures begins at about the same time as controls. SM50 transcripts continue to increase through the mesenchyme blastula stage, though at a lower level than controls. Spec 1 transcripts accumulate much less, attaining levels only slightly higher than the low level characteristic of the 12 h early blastula stage.

Fig. 4.

RNA blots showing the time course of SM50 (A) and Spec 1 (B) transcript accumulation in control and dissociated cultures. The various RNAs were extracted simultaneously from either an undisrupted, control culture (lanes a,b,c,d) or from a culture dissociated at the 4-cell stage and cultured in CFSW (lanes b′,c′,d′)- After the 4-cell stage, (lane a), samples were taken from both control and dissociated cultures when intact control embryos had reached the 7h midclcavage (b,b′), 12 h early blastula morula (c,c′) and 24 h mesenchyme blastula (d,d′) stages. Longer exposure of panel A (SM50) showed a slight accumulation of SM50 transcript in 12 h early blastulae (c,c′).

Fig. 4.

RNA blots showing the time course of SM50 (A) and Spec 1 (B) transcript accumulation in control and dissociated cultures. The various RNAs were extracted simultaneously from either an undisrupted, control culture (lanes a,b,c,d) or from a culture dissociated at the 4-cell stage and cultured in CFSW (lanes b′,c′,d′)- After the 4-cell stage, (lane a), samples were taken from both control and dissociated cultures when intact control embryos had reached the 7h midclcavage (b,b′), 12 h early blastula morula (c,c′) and 24 h mesenchyme blastula (d,d′) stages. Longer exposure of panel A (SM50) showed a slight accumulation of SM50 transcript in 12 h early blastulae (c,c′).

Fig. 5.

Quantification of SM50 (A) and Spec 1 (B) transcript accumulation. Data points were determined by densitométrie readings of RNA blot autoradiograms (similar to those shown in Fig. 4) plotted against time after fertilization. Dissociation was performed at the 4-cell stage and samples taken from control and dissociated cultures simultaneously at 7, 12, 18, and 24 h after fertilization. Graph A is from a blot hybridized with the SM50 probe while graph B is from an identical blot hybridized with the Spec 1 probe. All values are expressed as a percent of SM50 (A) or Spec 1 (B) transcript present in the normal mesenchyme blastula sample

Fig. 5.

Quantification of SM50 (A) and Spec 1 (B) transcript accumulation. Data points were determined by densitométrie readings of RNA blot autoradiograms (similar to those shown in Fig. 4) plotted against time after fertilization. Dissociation was performed at the 4-cell stage and samples taken from control and dissociated cultures simultaneously at 7, 12, 18, and 24 h after fertilization. Graph A is from a blot hybridized with the SM50 probe while graph B is from an identical blot hybridized with the Spec 1 probe. All values are expressed as a percent of SM50 (A) or Spec 1 (B) transcript present in the normal mesenchyme blastula sample

Dissociated cultures accumulated somewhat variable amounts of SM50, Spec 1 and Cy I transcript. One possible explanation for this variability may be that eggs from different females produce blastomeres which are differentially sensitive to the dissociation procedure. Often embryos from one egg batch fall apart quickly, while those from another batch cling more tenaciously during dissociation. It should be noted, however, that within any given experiment, relative SM50 and Cy I transcript levels were always greater than Spec 1 levels and this difference was statistically significant (P<0.05).

Expression of SM50 and Spec 1 occur in the appropriate cell types

To determine whether SM50 expression was lineage specific in dissociated cultures, embryos were dissociated at the 16-cell stage and blastomere types separated on sucrose gradients (Kitajima and Matsuda, 1982). Mesomeres and macromeres, blastomeres that produce progenitors of aboral ectoderm, are not easily separated from each other on these gradients, and they were cultured together in CFSW. Micromeres, the progenitors of the primary mesenchyme, were cultured separately, also in CFSW. RNAs prepared from such cultures were probed with both SM50 and Spec 1 (Fig. 6). SM50 transcript, characteristic of the micro-mere lineage (Benson et al. 1987), was present in both control mesenchyme blastula embryos and the dissociated micromere cultures which had developed for 24 h (Fig. 6A. lanes b,c) but was absent from both the initial 16-cell embryos and dissociated meso/macromere cultures which had developed for 24 h (lanes a,d). In contrast, Spec 1 transcript, characteristic of the aboral ectoderm, was present in mesenchyme blastula controls (6B,b) and at very low levels in dissociated meso/macromere cultures after 24 h (Fig. 6B, lane d) but Spec 1 was not detectable in either 16-cell-stage embryos (Fig. 6B,a) or cultures of dissociated micromeres after 24 h (Fig. 6B,c). These experiments suggest that in dissociated embryos transcription of these tissue-specific genes remains restricted to the appropriate cell types. It should be noted that the level of Spec 1 transcript was very low in the dissociated meso/macromere cultures. While this result is similar to the result obtained with dissociation at the 2-cell stage (see above), in this instance, the separation procedure at the 16-cell stage may have selectively reduced cell viability of the meso/macromeres. By 24 h after dissociation at the 16-cell stage, the meso/ macromere cultures contained between 20–50% undivided cells, (depending on the experiment) and cell lysis was apparent. In contrast, the smaller, and presumably hardier, micromeres appeared more robust under the same conditions.

Fig. 6.

RNA blot comparing the amounts of SM50 (A) and Spec 1 (B) transcript that accumulated in cultured, isolated cell types. Isolated micromeres (c lanes) were separated from meso/macromeres (d lanes) at the 16-cell stage and cultured as described in Materials and methods. RNAs from both 16-cell stage embryos (a lanes) and intact, control mesenchyme blastulae (b lanes) are included for comparison. The results from two separate experiments are shown here.

Fig. 6.

RNA blot comparing the amounts of SM50 (A) and Spec 1 (B) transcript that accumulated in cultured, isolated cell types. Isolated micromeres (c lanes) were separated from meso/macromeres (d lanes) at the 16-cell stage and cultured as described in Materials and methods. RNAs from both 16-cell stage embryos (a lanes) and intact, control mesenchyme blastulae (b lanes) are included for comparison. The results from two separate experiments are shown here.

Therefore, in order to evaluate possible differential susceptibility of different blastomeres to dissociation at the 16 cell stage, we performed in situ hybridization on smears of whole cells that had been dissociated at the 2cell stage and cultured in CFSW until the equivalent of the mesenchyme blastula stage. This procedure eliminates the preferential susceptibility of the larger meso/ macromeres to lysis or death. Since all cells are equivalent in size at the 2-cell stage, much less cleavage inhibition and cell death occur when embryos are dissociated at this stage. Since all cells in the dissociated cultures have nearly identical morphologies, micromere descendants can not be distinguished from mesomeres or macromere descendants. Therefore, in order to determine whether SM50 and Spec 1 were expressed in the appropriate cell types, we counted the number of positive cells in the dissociated culture and compared this to the number of cells that express these transcripts in control mesenchyme blastula embryos.

SM50 was expressed in approximately 5.5% of the cells from an intact mesenchyme blastula control culture (Fig. 7A,B). This percentage (5.5 %) is very close to the 6.4% predicted from a lineage analysis (Cameron et al. 1987), which indicates that there are 32 spicule-forming micromere descendents out of 500 total cells at the mesenchyme blastula stage. Dissociated 4-cell-stage embryos cultured to an equivalent of the mesenchyme blastula stage expressed SM50 in 4.4±1.9% (average of four experiments) of the total cell population. Grain counts per positive cell were similar in both control (26±7.7 grains/cell) and dissociated (25.1±5.4 grains/cell) cultures. This suggests that the number of cells expressing SM50, and not the amount of SM50 transcript per cell, is lower in dissociated than control cultures. The results clearly show that SM50 expression is restricted to a small number of cells and that the level of expression per cell is very close to the controls. This result, taken in conjunction with the culture of isolated micromeres, conclusively establishes that expression of SM50 in the micromere lineage does not require close-range cell interaction.

Fig. 7.

In situ hybridizations on single cells isolated from mesenchyme blastulae and on cultured, dissociated cells. Brightfield micrographs (A,C,E,G) and darkfield micrographs (B,D,F,H) of in situ hybridizations using [3H]labeled SM50 (A,B,C,D ) and Spec 1 (E,F,G,H) riboprobes. 5.5 % of the cells in the control culture (A,B) and 3.1 % of cells in this dissociated culture (C,D) accumulated SM50 transcripts. In contrast, Spec 1 was expressed in 31 % of cells from a control culture (E,F) and in 32% of the cells from a dissociated culture. Different autoradiographic exposures were used for G and H to optimize visual comparison in the figure; under identical conditions the cells expressing Spec 1 in dissociated cultures are much less heavily labeled than positive cells from intact embryos.

Fig. 7.

In situ hybridizations on single cells isolated from mesenchyme blastulae and on cultured, dissociated cells. Brightfield micrographs (A,C,E,G) and darkfield micrographs (B,D,F,H) of in situ hybridizations using [3H]labeled SM50 (A,B,C,D ) and Spec 1 (E,F,G,H) riboprobes. 5.5 % of the cells in the control culture (A,B) and 3.1 % of cells in this dissociated culture (C,D) accumulated SM50 transcripts. In contrast, Spec 1 was expressed in 31 % of cells from a control culture (E,F) and in 32% of the cells from a dissociated culture. Different autoradiographic exposures were used for G and H to optimize visual comparison in the figure; under identical conditions the cells expressing Spec 1 in dissociated cultures are much less heavily labeled than positive cells from intact embryos.

The in situ hybridization results with Spec 1 are also consistent with the lineage analysis of Cameron et al. (1987). 31% of the cells express Spec 1 in control embryos; the lineage analysis predicts that there are 200 aboral ectoderm cells of the approximate 500 cells present at the mesenchyme blastula stage (Fig. 7E,F). However, in contrast to SM50 expression, the low Spec 1 transcript levels in the dissociated cultures were extremely difficult to detect by this method. In three out of four experiments using 3H-probes, the numbers of grains per cell were only slightly above background levels and it was therefore not possible to determine the per cent of cells expressing Spec 1. In the fourth experiment (Fig. 7G,H), the grain density over cells was higher than in the three other experiments; background grain density over cells was 1.75±1.6 grains per cell and thus any cell with >6 grains (three fold higher than background) was considered positive. Using these criteria, 32% of the cells from the dissociated culture were scored positive for Spec 1. Additional experiments were carried out in which dissociated cultures were sectioned after embedding, rather than processed as smears. 35S-labeled Spec 1 probes were used rather than 3H ones. In three experiments, 23 %, 24 % and 31 % of the cells in dissociated cultures were labeled (twice background levels or greater). Hence, even though Spec 1 expression is very near the limit of detectability, the results with in situ hybridization show that a subset of the population is expressing the gene at low levels, consistent with the proposition that the transcript accumulation is most likely restricted to the appropriate cell type.

Our use of tissue-specific markers and dissociation is an thereby minimizing close-range cell interactions. This son, 1925; Davidson, 1986) in which blastomeres are removed from embryos and cultured in isolation, extension of classical embryological experiments (Wil-type of experiment provides information about the possible dependence of steps in the pathway on cell interactions. Until the recent development of suitable molecular markers, the outcome of such experiments had to be judged by the appearance of tissue-specific characters which almost always require morphogenetic movements and aggregations of many cells. Earlier studies of development in sea urchins by Giudice and his colleagues (1962, 1970), as well as by Hynes and Gross (1972) and Spiegel and Spiegel (1986) had no available tissue- and lineage-specific markers. Arceci and Gross (1980) did show that the switch from early to late histone variant synthesis did not require homotypic cell contact. Some recent studies on amphibian embryos show a very wide range of results using different tissue-specific molecular markers. In some instances, cell associations were not required (Sargent et al. 1986); in other instances lack of Ca2+ and Mg2+ could be tolerated, but not cell dispersal (Sargent et al. 1986; Gurdon et al. 1984). Yet other markers were completely prevented in their expression by dissociation (Jones and Woodland, 1986).

The results obtained here show that expression of two very different markers, SM50 (characteristic of the micromere lineage) and Spec 1 (characteristic of the aboral ectoderm tissue), are both expressed autonomously in the absence of close-range cell interactions. Both initiate zygotic accumulation at near the normal time and both accumulate in a subset of the total disaggregated population. The evidence from cultures of partially purified cell types, and from in situ hybridizations supports the idea that correct cells are expressing these markers. Perhaps it is not surprising that SM50 is expressed autonomously because it is known that isolated micromere cultures will construct spicules (Okazaki, 1975). However, the present experiments show that there is not even a requirement for homotypic interactions in this lineage. The correct time and place of expression of Spec 1 demonstrates that neither closerange interactions nor position (since this is randomized in stirring cultures of disaggregated cells) influence expression of this gene characteristic of ectodermal tissue. It is likely, therefore, that the initial expression of both these genes is autonomous, and is regulated at least in part by factors present in the uncleaved zygote. These experiments do not rule out the importance of freely diffusible stable factors, but since the response to any such postulated factors is still specific to subsets of the whole population, the crucial ability of cells to regulate temporal and spatial gene expression must still be autonomous.

There is a difference, and an important one, between SM50 and Spec 1 expression. While SM50 expression is diminished somewhat in dissociated cultures, it is generally expressed at least at half the normal level. In situ hybridization studies showed that the observed quantitative reduction is principally due to a somewhat smaller number of cells accumulating the transcripts at near normal levels. Spec 1 expression, on the other hand, is quantitatively reduced very extensively. Although expression starts near the normal time, continued accumulation is severely depressed. Hurley et al. (1989) have recently observed this same quantitative depression with several markers expressed in aboral ectoderm. Reduction of Spec 1 expression is probably due to a substantially reduced RNA accumulation in a subset of the total population of dissociated cells. It is difficult to be confident about the precise number of cells in dissociated cultures that actually do express Spec 1 because of the considerable technical difficulty of detecting such low transcript levels by in situ hybridization. Our results clearly could have detected a few cells expressing Spec 1 at normal levels, but we do not observe this.

Clearly the response to dissociation and culture in CFSW is gene specific and there are several possible explanations for this observation. Cells cultured in CFSW terminate rapid cleavage cell divisions just a little earlier than controls (cf. Fig. 2) and it is possible that Spec 1 accumulation could be regulated by cell cycle timing mechanisms, as is the synthesis of cholinesterase in ascidian embryos (Satoh and Ikegami, 1981). However, we have recently cultured embryos in CFSW but allowed the cells to remain closely associated within the fertilization membrane. Under these conditions, cell division behaviour is the same as in stirred, dissociated cultures but Spec 1 is expressed at near normal levels (unpublished results). Hence, neither failure to complete all the cell divisions, nor, indeed, the lack of Ca, seem to be responsible for the low level of Spec 1 accumulation in dissociated cultures. Hurley et al. (1989) reached a similar conclusion on the basis of experiments in which addition of Ca2+ to rapidly stirred dissociated cultures did not enhance Spec 1 expression. An alternative explanation for the lowered levels of Spec 1 transcript in dissociated cultures is that continued accumulation of this transcript is dependent on some form of cell interaction. While we have not found any evidence to contradict this hypothesis, further experiments are necessary before any firm conclusions can be drawn.

In summary, we have demonstrated that presumptive mesenchymal and ectodermal cells retain the ability to accumulate tissue-specific markers in the absence of close-range cell interactions. The temporal and spatial fidelity of both SM50 and Spec 1 transcript accumulation is remarkable and suggestive of localized regulative factors present in the uncleaved zygote. Statistically significant, quantitative differences exist between SM50 and Spec 1 transcript accumulation in dissociated cultures. Whether diminished accumulation of Spec 1 transcript is actually due to the lack of cell contact is currently under investigation.

We greatly appreciate the contributions of George Shifflett and Dick Poccia who gave generously of their time and expertise. The suggestions from Steve Benson, Nick George, Chris Killian and John Gerhart are also gratefully acknowledged. We are thankful for the information and criticisms shared with us by David Hurley, Lynn Angerer and Robert Angerer. We are also grateful to Doug Melton, Bill Klein, Andrew Cameron, and Lynn and Robert Angerer for gifts of cloned DNA. This work was supported by a grant from NIH no. HD 15043. Gene expression in dissociated embryo cells 307

Angerer
,
R. C.
&
Davidson
,
E. H.
(
1986
).
Molecular indices of cell lineage specification in sea urchin embryos
.
Science
226
,
1153
1160
.
Angerer
,
L. M.
,
Soler
,
M. H.
&
Angerer
,
R. C.
(
1987
).
In Situ hybidization with RNA probes - an anotated recipe
. In
In Situ Hybridization: Application to Neurobiology
(ed.
K. L.
Valentino
,
J. H.
Eberwine
, &
J. D.
Barchas
), pp.
42
-
72
.
Oxford University Press
,
New York
.
Arceci
,
R. J.
&
Gross
,
P. R.
(
1980
).
Histone gene expression: Progeny of isolated early blastomeres in culture make the same changes as in the embryo
.
Science
209
,
607
609
.
Benson
,
S.
,
Sucov
,
H.
,
Stephens
,
L.
,
Davidson
,
E.
&
Wilt
,
F.
(
1987
).
A lineage-specific gene encoding a major matrix protein of the sea urchin embryo spicule
.
Devi Biol
.
120
,
507
519
.
Cameron
,
R. A.
,
Hough-Evans
,
B. R.
,
Britten
,
R. J.
&
Davidson
,
E. H.
(
1987
).
Lineage and fate of each blastomere of the eight-cell sea urchin embryo
.
Genes and Dev
.
1
,
75
84
.
Carpenter
,
C. D.
,
Bruskin
,
A. M.
,
Hardin
,
P. E.
,
Keast
,
M. J.
,
Anstrom
J.
,
Tyner
,
A. L.
,
Brandhorst
,
B. P.
&
Klein
,
W. H.
(
1984
).
Novel proteins belonging to the troponin C superfamily are encoded by a set of mRNAs in sea urchin embryos
.
Cell
36
,
663
671
.
Cox
,
K. H.
,
Angerer
,
L. M.
,
Lee
,
J. J.
,
Davidson
,
E. H.
&
Angerer
,
R. C.
(
1986
).
Cell lineage-specific programs of expression of multiple actin genes during sea urchin embryogenesis
.
J. molec. Biol
.
188
,
159
172
.
Cox
,
K. H.
,
Deleon
,
D. V.
,
Angerer
,
L. M.
&
Angerer
,
R. C.
(
1984
).
Detection of mRNAs in sea urchin embryos by in Situ hybridization using asymmetric RNA probes
.
Devi Biol
.
101
,
485
502
.
Davidson
,
E. H.
(
1986
).
Gene Activity in Early Development, Third edition
.
Academic Press, Inc
.,
New York
.
Davidson
,
E.
(
1989
).
Lineage-specific gene expression and the regulative capacities of the sea urchin embryo: a proposed mechanism
.
Development
105
,
421
445
.
Driesch
,
H.
(
1891
).
Entwicklungsmechanische studien
.
Z. Wzss. Zool
.
53
,
160
184
.
Eldon
,
E. D.
,
Angerer
,
L. M.
,
Angerer
,
R. C.
&
Klein
,
W. H.
(
1987
).
Spec 3: Embryonic expression of a sea urchin gene whose product is involved in ectodermal ciliogenesis
.
Genes and Dev
.
1
,
1280
1292
.
Giudice
,
G.
(
1962
).
Restitution of whole larvae from dissagregated cells of sea urchin embryos
.
Devi Biol
.
5
,
402
.
Giudice
,
G.
(
1986
).
The Sea Urchin Embryo: A Developmental Biological System. SpringerVerlag, Berlin, Heidelberg, New York, Tokyo
.
Giudice
,
G.
&
Mutolo
,
V.
(
1970
).
Reaggregation of dissociated cells of sea urchin embryos
.
Adv. Morphogen
.
8
,
115
.
Gurdon
,
I. B.
(
1987
).
Embryonic induction - molecular prospects
.
Development
99
,
285
306
.
Gurdon
,
J. B.
,
Brennan
,
S.
,
Fairman
,
S.
&
Mohun
,
T. J.
(
1984
).
Transcription of muscle-specific actin genes in early xenopus developmentmuclear transplantation and cell dissociation
.
Cell
38
,
691
700
.
Hall
,
H. G.
(
1978
).
Hardening of the sea urchin fertilization envelope by peroxidase-catalyzed phenolic coupling of tyrosine
.
Cell
15
,
343
355
.
Hinegardner
,
R.
(
1967
).
Echinoderms
. In
Methods in Developmental Biology
(ed.
F.
Wilt
and
N. K.
Wessels
), pp.
139
155
.
T. Y. Crowell
,
New York
.
HORSTADIUS
, (
1973
).
Experimental Embryology of Echinoderms. Clarendon, Oxford
.
Hurley
,
D. L
,
Angerer
,
L. M.
&
Angerer
,
R. C.
(
1989
).
Altered expression of spatially regulated embryonic genes in the progeny of separated sea urchin blastomeres
.
Development
106
,
567
579
.
Hynes
,
R. O.
,
Raff
,
R. A.
&
Gross
,
P. R.
(
1972
).
Properties of the three cell types in 16 cell sea urchin embryos: aggregation and microtubule protein synthesis
.
Devi Biol
.
27
,
150
164
.
Jones
,
E. A.
&
Woodland
,
H. R.
(
1986
).
Development of the ectoderm in Xenopus: Tissue specification and the role of cell association and division
.
Cell
44
,
345
355
.
Killian
,
C. E.
&
Wilt
,
F. H.
(
1989
).
The accumulation and translation of a spicule matrix protein mRNA during sea urchin development
.
Devi Biol
.
133
,
148
156
.
Kitajima
,
T.
&
Matsuda
,
R.
(
1982
).
Specific protein synthesis of sea urchin micromeres during differentiation
.
Zool. Mag
.
91
,
200
205
.
Lee
,
J.
,
Shott
,
R.
,
Rose
,
S.
,
Thomas
,
T.
,
Britten
,
R.
&
Davidson
,
E.
(
1984
).
Sea urchin actin gene subtypes
.
J. molec. Biol
.
172
,
149
176
.
Maniatis
,
T.
,
Fritsch
,
E. F.
&
Sambrook
,
J.
(
1982
).
Molecular Cloning: A Laboratory Manual, pp. 202-203. Cold Spring Harbor, New York
.
March
,
C. J.
,
Mosley
,
B.
,
Carsen
,
A.
,
Cerretti
,
D. P.
,
Braedt
,
G.
,
Price
,
V.
,
Gillis
,
S.
,
Henney
,
C. S.
,
Kronheim
,
S. R.
,
Grabstein
,
K.
,
Conlon
,
P. J.
,
Hopp
,
T. P.
&
Cosman
,
D.
(
1985
).
Cloning, sequencing, and expression of two distinct human interleukin-1 complementary DNAs
.
Nature, Lond
.
315
,
641
647
.
McClay
,
David R.
&
Wessel
,
Gary M.
(
1985
).
The surface of the sea urchin embryo at gastrulation: a molecular mosaic
.
TIGS
1
,
12
16
.
Melton
,
D.
,
Krieg
,
P. A.
,
Rebagliati
,
M. R.
,
Zinn
,
K.
&
Green
,
M. R.
(
1984
).
Efficient in vitro synthesis of biologically active RNA and RNA hybridization probes from plasmids containing a SP6 bacteriophage promoter
.
Nucl. Acids Res
.
12
,
7035
7056
.
Okazaki
,
K.
(
1975
).
Spicule formation by isolated micromeres of the sea urchin embryos
.
Amer. Zool
.
15
,
567
581
.
Sargent
,
T. D.
,
Jamrich
,
M.
&
Dawid
,
I. B.
(
1986
).
Cell interactions and the control of gene activity during early development of Xenopus laevis
.
Devi Biol
.
114
,
238
246
.
Satoh
,
N.
&
Ikegami
,
S.
(
1981
).
A definite number of aphidicolin-sensitive cell-cycle events are required for acetylcholinesterase development in the presumptive muscle cells of the ascidian embryo
.
J. Embryol. exp. Morph
.
64
,
1
13
.
Schroeder
,
T. E.
(ed.) (
1986
).
Echinoderm gametes and embryos
. In
Methods in Cell Biology
27
.
Academic Press
:
New York
. pg.
470
.
Spiegel
,
E.
&
Spiegel
,
M.
(
1986
).
Cell-cell interactions during sea urchin morphogenesis
.
In Developmental Biology, The Cellular Basis of Morphogenesis
, vol.
2
(ed.
L. W.
Browder
), pp.
195
240
.
New York
:
Plenum Press
.
Sucov
,
H. N.
,
Benson
,
S.
,
Robinson
,
J. J.
,
Britton
,
R. J.
,
Wilt
,
F.
&
Davidson
,
E. H.
(
1987
).
A lineage-specific gene encoding a major matrix protein of the sea urchin embryo spicule
.
Devi Biol
.
120
,
507
519
.
Thomas
,
P. S.
(
1980
).
Hybridization of denatured RNA and small DNA fragments transferred to nitrocellulose
.
Proc. natn. Acad. Sci. U.S.A
.
77
,
5201
5205
.
Wells
,
D. E.
,
Bruskin
,
A. M.
,
O’Brochta
,
D. A.
&
Raff
,
R. A.
(
1982
).
Prevalent RNA sequences of mitochondrial origin in sea urchin embryos
.
Devi Biol
.
92
,
557
562
.
Wilson
,
E. B.
(
1925
).
The Cell in Development and Heredity, Third Edition
.
Macmillan, New York
.
Wilt
,
F. H.
(
1987
).
Determination and morphogenesis in the sea urchin embryo
.
Development
100
,
559
575
.