Lineage segregation and developmental autonomy in expression of functional muscle acetylcholinesterase mRNA in the ascidian embryo

DEVELOPMENTAL

BIOLOGY

105, 479-487

(1984)

Lineage Segregation and Developmental Autonomy in Expression of Functional Muscle Acetylcholinesterase mRNA in the Ascidian Embryo H.

THOMAS Boston

University

Marine

Program,

Received

April

MEEDEL Marine

AND Biological

4, 1984; accepted

J. R.

WHITTAKER

Laboratory, in revised

Woods Hole, form

June

Massachusetts

02543

6, 1984

Acetylcholinesterase is a histospecific marker of cell differentiation occurring only in the muscle and mesenchyme tissues of the ascidian embryo. The distribution of functional mRNA coding for this enzyme has been investigated and it is shown here that only cells of muscle and mesenchyme lineages possess such a template. Blastomeres of four cell lineage quadrants were separated microsurgically from eight-cell-stage embryos of Ciona intestina1i.s and raised in isolation until muscle development was well advanced. Measurement of enzyme activity in the resulting partial embryos revealed that acetylcholinesterase was limited to descendants of one blastomere pair, the B4.1 blastomeres containing muscle and mesenchyme lineages. To study the tissue distribution of acetylcholinesterase mRNA, RNA from partial embryos was translated in Xenqpus lueuti oocytes. When oocytes were injected with an appropriate template, they synthesized a biologically active acetylcholinesterase that could he selectively immunopurified with an antiserum to the ascidian enzyme. Under the conditions used the quantity of acetylcholinesterase mRNA was directly related to the enzyme activity in immunoprecipitates. Acetylcholinesterase mRNA was found only in B4.1 lineage partial embryos where it occurred in approximately the same amount as in whole embryos of the same age. Since there is a limited period from gastrulation until the middle tail-formation stage when functional acetylcholinesterase mRNA accumulates, the results of our mRNA distribution experiments strongly suggest that the gene for ascidian acetylcholinesterase is active only in muscle and mesenchyme tissues. The histospecific occurrence of this enzyme apparently does not involve selective, cell-specific control of translation.

INTRODUCTION

developing muscle and mesenchyme tissues of the larva (Durante, 1956; Meedel and Whittaker, 1979). Since the enzyme subunit is probably a single polypeptide, enzyme development conceivably results from the expression of a single structural gene (Meedel, 1980). We have shown recently by direct translational assay in Xenopus oocytes that a functional acetylcholinesterase mRNA first occurs at gastrulation and reaches a plateau of concentration by the middle tail-formation stage (Meede1 and Whittaker, 1983). This finding and a coincident window of stage-specific sensitivity to actinomycin D, an inhibitor of RNA synthesis (Meedel and Whittaker, 1979), imply that transcription of the acetylcholinesterase gene first begins at gastrulation. In the ascidian embryo the acetylcholinesterase gene, and muscle development in general, seem clearly to be under a control mediated by egg cytoplasmic substances. This conclusion is evident from classic studies on cell lineage and the mosaic development of partial embryos (Whittaker, 1979), and is confirmed by recent findings that muscle cell lineage cytoplasm can initiate acetylcholinesterase development when it is moved to cells of nonmuscle lineages (Whittaker, 1980, 1982). While our previous results indicate that the factor is not a segregated maternal mRNA for the enzyme (Meedel and Whittaker, 1983), such control might possibly

At present, perhaps the single most intriguing problem in developmental biology is how selective gene expression is regionally controlled in the embryo (Gurdon, 1981). Answers to such questions as when certain genes are first expressed during embryonic development and which cells in the embryo express these genes become a necessary and important prelude to testing ideas about the nature of the mechanisms involved in regional gene control. In the so-called mosaic embryos of many invertebrates, an early determination of cell fate occurs, with restricted “cell lineages” leading to later regional differentiations of the embryo. A presumption in such work is that egg cytoplasmic factors and their early positioning determine the cell lineages of the embryo, and are possibly themselves the agents which later control selective gene expression (Davidson, 1976). One practical approach to problems of selective gene expression in embryos is to seek the functional messenger RNAs of genes that code for histospecific proteins of known and identifiable biological activity. An acetylcholinesterase of the ascidian embryo has proven to be a particularly useful marker of muscle cell differentiation because during premetamorphic development its appearance is confined exclusively to the 479

0012-1606/84 Copyright All rights

$3.00

0 1984 by Academic Press, Inc. of reproduction in any form reserved.

480

DEVELOPMENTAL

BIOLOGY

operate by translational selectivity, as suggested for early molluscan development by Collier and McCarthy (1981). Translation rather than transcription may be regionally controlled in the mosaic embryos of invertebrates. If so, ascidian muscle genes must also function in the nonmuscle cell lineages. Our present findings with acetylcholinesterase mRNA distribution show that determinants do not operate against any such background of uniform gene expression. Activity of the muscle acetylcholinesterase gene begins at early gastrulation and follows a selective and highly autonomous expression in muscle lineage cells of the embryo. Based on these results, we propose that histospecific transcriptional control of genes rather than differential translational control of widely expressed mRNAs for the muscle proteins regulates muscle differentiation in ascidian embryos. MATERIALS

AND

METHODS

Embryos. Adult Ciona intestinalis (L.) (subphylum Urochordata, class Ascidiacea) collected in the vicinity of Woods Hole, Massachusetts, were maintained on sea tables with continuously flowing seawater and constant light. Gametes were obtained surgically from the gonoducts; embryos were raised at 18 f 0.1% in sterile Millipore-filtered (0.2-pm porosity) seawater. For some purposes fertilized eggs were dechorionated manually with sharpened steel needles. These embryos were cultured afterward in sterile Millipore-filtered seawater containing 0.1 mM EDTA (Crowther and Whittaker, 1983). Partial embryo isolations. At the eight-cell stage (Fig. 1) embryos were separated with agar-coated glass filament needles into animal half-embryos (the a4.2 and b4.2 cell pairs) and vegetal half-embryos (the A4.1 and B4.1 cell pairs). The vegetal quartets were divided further into A4.1 and B4.1 blastomere pairs. Three kinds of partial embryo were thus obtained: animal quartet, A4.1 pair, and B4.1 pair. They were reared together as partial embryos in the same agar-coated Syracuse dish containing sterile Millipore-filtered seawater with EDTA. Some dechorionated whole embryos were included with the partial embryos in the dish to serve as controls in various of the analytical procedures used with the partial embryos. Surgical techniques and embryo orientation features used in selecting the appropriate cells are described elsewhere (Whittaker et al, 19’77; Whittaker, 1982). Pur$ication of RNA. RNA was prepared from embryos of various developmental stages by extraction with phenol and chloroform (Meedel and Whittaker, 19’78). Any remaining protein was digested with proteinase K and the resulting RNA was reextracted with

VOLUME

105, 1984

phenol and chloroform; tRNA, DNA, and polysaccharides were removed by washing the RNA with sodium acetate (Lee et al, 1978). Purified RNA was dissolved in water and stored at -85°C until used. RNA from all developmental stages directed the synthesis of high-molecular-weight polypeptides in a mRNA-dependent reticulocyte lysate (Pelham and Jackson, 1976), and was therefore judged to be relatively intact. Because of the limited amount of material available, a scaled down version of the above procedure was used to purify RNA for mRNA distribution studies. Groups of 25 dechorionated embryos or partial embryos were digested in 25 ~1 of buffer (10 mM Tris-HCl pH 7.6, 5 mMEDTA, 1% SDS) containing 100 fig/ml of proteinase K. After 90 min at 37”C, NaCl was added to a concentration of 100 mM, and the samples were extracted with phenol-chloroform. The aqueous phase was removed after centrifugation (15009 for 10 min at 15”C), and the organic phase was reextracted with 25 ~1 of buffer containing 100 mM NaCl, and recentrifuged. The combined aqueous phases were ethanol precipitated, dissolved in water, and stored at -85°C. Recovery, monitored by adding radioactive morula and gastrula stage RNA from Lytechinus pictu.s (gift of Joan Ruderman) to the initial digestion mixture, was 85-95s. Acetylcholinesterase localizaticm Our usual fixation procedure in which embryos are incubated for 3 min in cold (5”C), 80% ethanol (Durante, 1956) did not yield reproducible histochemical staining of acetylcholinesterase in whole dechorionated or partial embryos. This difficulty was overcome by fixation for 10 min at 5°C in Karnovsky’s solution (1965) containing one-half the suggested concentrations of glutaraldehyde and formaldehyde. Fixed embryos were washed for 10 min in the recommended buffer (Karnovsky, 1965) and reacted for 3 hr (18°C) for acetylcholinesterase activity (Karnovsky and Roots, 1964). Because the incipient formation of epidermal test material in 12-hr control embryos restricts the passage of histochemical reagents slightly, the reaction in whole embryos was carried out for 6 hr. Stained embryos were dehydrated in ethanol, cleared in xylene, and mounted in Damar resin for microscopic examination and photography. Acetylcholinesterase assay Quantitative measurement of acetylcholinesterase was done radiometrically by the method of Schrier et al. (1974). Dechorionated embryos and partial embryos were cytolyzed in 40 ~1 of assay buffer (62.5 mM potassium phosphate pH 7.4, 1.25 mM EDTA, 250 mM NaCl, 0.625% Triton X-100). Ten microliters of [acetyZ-l-‘4C]acetylcholine iodide (New England Nuclear, 1.7 mCi/mmole) in water were added to give a final concentration of 0.9 mM, and the samples were incubated at 18°C in a constant temperature water bath for the times stated.

MEEDEL

AND WHITTAKER

Lineage-Spe&c

Acetylcholinesterase purified from injected oocytes was assayed in a similar manner; however, the insolubility of immunoprecipitates required that such samples be incubated with constant agitation. Also, because the level of enzyme activity was quite low, a higherspecific activity substrate, [acet&lJ4C]acetylcholine chloride (Amersham, 55.3 mCi/mmole; final concentration 0.1 mM) was used in the mRNA distribution experiments. With larval acetylcholinesterase, enzyme activity measured under the above conditions was linear for at least 40 hr if less than 25-30s of the substrate was hydrolyzed. The spontaneous hydrolysis of acetylcholine in assay buffer during the period of enzyme assay has been subtracted from all reported data. Oocyte microinjection and acetylcholinesterase jication. Female Xenopua laevis were purchased

puri-

from NASCO (Fort Atkinson, Wise.). Stages five and six oocytes (Dumont, 1972) were dissected free of smaller oocytes and stored at 18°C in modified Barth’s medium (Gurdon, 1977). RNA was injected into oocytes in a volume of 40 nl; control oocytes were injected with the same volume of either water or carrier RNA. A modification of the device described by Tommerup (1982) was used for all injections; this method was reproducibly accurate to within 3-5s when volumes of 20 nl or greater were injected. Up to 50 oocytes were placed in plastic tissue culture dishes containing 5 ml of modified Barth’s medium and incubated at 18°C. Control experiments showed that when 50 ng of RNA from 11-hr embryos was injected, essentially no ascidian acetylcholinesterase was found in the culture medium until after 36-48 hr of incubation. At specified times, groups of oocytes were removed from the incubation medium into 1.5 ml polypropylene centrifuge tubes, crushed with a glass stirring rod, and extracted with 50 mM Tris-HCl, pH 7.2,145 mM NaCl, 2% Triton X-100, 1% sodium deoxycholate. After 30 min on ice with intermittent stirring, oocyte homogenates were centrifuged at 12,000~ for 5 min and the supernatant between the floating lipid layer and the pigmented residue was removed. The remaining material was reextracted and the two supernatants were combined. Acetylcholinesterase was purified by adding a Ciona acetylcholinesterase antiserum (Meedel, 1980) and using Formalin-fixed Staphylococcus aureus (Immuno-Precipitin, Bethesda Research Laboratories) to recover enzyme-antibody complexes (Kessler, 1975). The insoluble complexes (“immunoprecipitates”) were washed three times with 50 mM Tris-HCl, pH 7.2, 145 mM NaCl, 0.5% Triton X-100 and assayed for acetylcholinesterase activity. Immunoprecipitates prepared in this manner specifically retain only Ciona acetylcholinesterase (Meedel and Whittaker, 1983). All experi-

Gene

481

Expression.

ments reported in the present communication were done under conditions that resulted in linear acetylcholinesterase production with respect to the amount of RNA injected per oocyte and to the length of time that oocytes were incubated. RESULTS

Acetylcholinesterase Is Expressed cmly in Descendants of the B4.1 Lineage

Microdissections were done on embryos at the eightcell stage when the major lineages of the ascidian larva have either partially or completely sorted out (Conklin, 1905; Ortolani, 1955). During normal development, the B4.1 blastomere pair gives rise to most of the muscle and mesenchymal tissues, as well as to part of the endoderm, whereas the animal-half descendants (a4.2 and b4.2) become epidermis, brain, sense organs, and palps; progeny of the A4.1 yield endoderm, notochord, and brain (reviewed in Reverberi, 1971). In addition to having at least partially segregated the major larval lineages, eight-cell-stage embryos offer technical advantages for microsurgery because the cells are still large enough to make such operations relatively simple. The unique cleavage pattern formed at this stage (Fig. 1) also permits isolated blastomeres to be identified with certainty. Acetylcholinesterase is found only in the tail muscle and mesenchyme tissues of postneurula stage ascidian larvae (Meedel and Whittaker, 1979) and, therefore, only descendants of the B4.1 blastomere pair should produce this enzyme. Previous studies have verified this expectation (Whittaker et aL, 1977; Whittaker, 1982). We have repeated these experiments using embryos and partial embryos cultured until the middle

ANIMAL

b 4.2

ANTERIOR

n

POSTERIOR

84.1

VEGETAL FIG. 1. Lateral-view diagram of cytoplasmic territories in the bilaterally symmetrical ascidian eight-cell stage as mapped by Conklin (1905) and Ortolani (1954). Nomenclature of the blastomere pairs is from Conklin (1905). Abbreviations are muscle (m), mesenchyme (mch), nervous system (n), and notochord (ch). The two polar bodies are shown resting on an a4.2 blastomere.

482

DEVELOPMENTAL

BIOLOGY

tail-formation stage, 12 hr after fertilization. As expected, acetylcholinesterase was detected histochemitally in the tail muscle of control, dechorionated embryos (Fig. 2) and in progeny derived from the B4.1 blastomere pair (Fig. 5), but was not found at this time in descendants of either A4.1 (Fig. 4) or animalhalf isolates (Fig. 3). Since B4.1 blastomeres contain territories destined to become endoderm (Ortolani, 1954) as well as muscle and mesenchyme, not all of their progeny cells produce acetylcholinesterase (Fig. 5). Some histochemical assays were done with partial embryos of the above kinds reared to hatching age (18-20 hr) before fixation. Enzyme activity was still not observed in either animal-half partial embryos or A4.1 partial embryos of this later age; incubation times longer than 3 hr also did not reveal localized staining but resulted only in darker background staining. Quarter-embryos derived from a4.2 or b4.2 blastomere pairs did not behave differently than the half-embryos originating from the combined blastomere pairs: none of these produced a localized acetylcholinesterase activity.

VOL~JME

105, 1984

Quantitative determination of acetylcholinesterase activity by radiometric assay (Schrier et al, 1974) confirmed the histochemical findings (Table 1). These measurements also showed that under favorable conditions, isolates derived from B4.1 blastomeres produced nearly the same level of enzyme activity as did whole control embryos. In three experiments this value was 80-90s of controls, and in a fourth it was slightly greater than the control. Perhaps some of this 10-20s disparity is caused by stresses resulting from the isolation procedure. We cannot at the moment, however, rule out the possibility that other minor sources of acetylcholinesterase derive from non-B4.1 lineages (see Discussion). In any event, our results demonstrated that B4.1 blastomeres reared in isolation behave quite normally with respect to acetylcholinesterase expression up to 12 hr of development. This makes these partial embryos a good model for studying certain aspects of muscle differentiation in ascidians. Also, despite the absence of other tissue lineages with which muscle cells might normally interact (see Fig. 5), the nearly quantitative

03

FIGS. 2-5. Histochemical localization of acetylcholinesterase activity in 1%hr Ciona embryos and partial embryos of the same age. Figure 2, dechorionated whole embryo (middle tail-formation stage). Figure 3, partial embryo from an animal-half quartet. Figure 4, partial embryo from an A4.1 blastomere pair. Figure 5, partial embryo from a B4.1 blastomere pair. All photographs are the same magnification; the bar in Fig. 5 is 50 pm.

MEEDEL

TABLE DISTRIBIJTION

OF ACETYLCHOLINESTERASE

AND

Lineage-Spec$c

WHITTAKER

1 IN 12-hr

Ciona

EMBRYOS

Series Acetylcholinesterase source

1

2

3

4

Whole embryo Animal half-embryo A4.1 quarter-embryo B4.1 quarter-embryo

7.12 0.07 0.02 5.95

5.48 0.08 0.08 5.70

8.38 0.11 0.13 7.34

7.27 0.08 0.04 5.80

0.84

1.04

0.88

0.80

B4.1: whole

embryo

ratio

Note. For each series, embryos were obtained from the fertilized eggs of a single animal. Samples consisting of 5-10 embryos or partial embryos were cytolyzed in buffer and the entire sample was assayed for periods of 6-24 hr to determine acetylcholinesterase activity. Results are expressed as micromoles (X105) acetylcholine hydrolyzed per embryo per hour.

regulation of acetylcholinesterase activity in isolated B4.1-derived partial embryos provides strong support for the idea that agents controlling larval muscle development are functionally autonomous. Up to the neural plate stage, 7 hr after fertilization, a low and unchanging level of acetylcholinesterase exists in Ciona (Meedel and Whittaker, 1979; Meedel, 1983). When all four blastomere pairs were isolated at the eight-cell stage and assayed immediately by the radiometric method, an essentially uniform distribution of this acetylcholinesterase activity was observed (data not shown). The significance of this activity is unknown. Because it was present in amounts similar to that seen in descendants of nonmuscle lineage blastomeres (i.e., progeny of animal-half and A4.1), it undoubtedly accounts for the low level of activity detected in these tissues (Table 1). Our earlier work with protein synthesis inhibitors showed that acetylcholinesterase activity of pregastrular embryos is quite stable, persisting until at least the middle tail-formation stage (Meedel and Whittaker, 1979). We suggest, therefore, that the acetylcholinesterase present in animal-half and A4.1 partial embryos is the preexistent enzyme found in the egg and is not governed by mechanisms relevant to the large increase in enzyme activity observed in the muscle lineage. Timing

of Acetylcholinesterase

mRNA

Accumulation

Previous studies of acetylcholinesterase expression in ascidians depended on the effects of metabolic inhibitors and their results suggested that transcription of the muscle acetylcholinesterase gene(s) occurs between gastrulation and the middle tail-formation stages (Whittaker, 1973; Meedel and Whittaker, 1979; Meedel, 1983). A direct method was developed for assaying

Gene

483

Expression

translatable acetylcholinesterase mRNA that uses microinjected X laevis oocytes as an in vivo protein synthesis system and specific antiserum to the ascidian enzyme (Meedel and Whittaker, 1983). Results obtained with this assay supported the conclusions from the inhibitor studies. With this same methodology, we now present a more extensive analysis of translationally active acetylcholinesterase mRNA in ascidian embryos that reveals some significant new features of the development of this template (Fig. 6). This more detailed study indicated a very low level of acetylcholinesterase mRNA in unfertilized eggs and 4-hr embryos (inset, Fig. 6). The amount detected in these early stages was several hundred times less than that found at 11 hr of development. Control experiments demonstrated that this mRNA behaved as expected for a template mRNA: the level of functional enzyme produced depended on both the quantity of RNA injected and the duration of oocyte incubation. The level of mRNA activity was always slightly higher in 4-hr embryos than in unfertilized eggs. Although this rise was diagrammed as a linear increase, its kinetics are not known. Quite possibly the increase in template begins at about 4 hr of development. Measurement of template activity in RNA prepared from embryos at hourly intervals during the 4-7 hr after fertilization showed the first significant increase in acetylcholinesterase mRNA at between 4-5 hr of development. This observation confirms our previous speculation, based on fewer data, that quantitatively significant production of functional mRNA for acetylcholinesterase probably begins at early gastrulation (Meedel and Whittaker, 1983). After the initial rapid rise, template activity continued to increase dramatically for about 6 hr, until the middle tail-formation stage when it was highest. Between middle and late tail-formation stages a slight decline in activity was noted. Perry and Melton (1983) obtained generally similar data for the occurrence of functional acetylcholinesterase mRNA by microinjecting polyadenylated RNA from ascidian embryos at various stages of development into Xenopus oocytes. Although these authors did not use an antiserum to identify oocyte translation products, only polyadenylated RNA from postgastrula stage embryos was found to elicit synthesis of acetylcholinesterase above the endogenous background; polyadenylated RNA from earlier stages did not. Distribution of Acetylcholinesterase Developing Partial Embryos

mRNA

in

The above results lend support to the idea of a rapid appearance during development of translatable acetyl-

484

DEVELOPMENTAL BIOLOGY

0

1 2

I8-cell3 14 I 32cell

I fertilization

5 6

I I

lote

early

VOLUME 105, 1984

7 8 9 10 11 12 13 14 15 16 17 18 hr

;leurol/

plate

eorly toil formation

I

mlddle tall formation

gostrula

I

late tail formation first hatching

gostrula

EMBRYONIC STAGES FIG. 6. Accumulation of translatable acetylcholinesterase mRNA in developing Ckmo embryos. RNA was prepared from embryos at the stages indicated and injected into Xenop~ oocytes. Incubation was for 16 hr and immunoprecipitated enzyme was assayed for 16 hr. Oocytes were injected with 50 ng of RNA; each point represents the average from three groups of eight oocytes. Imet. Oocytes were injected with 100 ng of RNA and incubated for 18 hr. Immunoprecipitated acetylcholinesterase was assayed for 84 hr; the bars represent standard deviation, calculated from four groups of eight oocytes injected for each point. The dotted lines represent standard deviations in four groups of eight water-injected oocytes.

cholinesterase mRNA; they indicate nothing, however, about the distribution of this template in the embryo. Given the apparent 2-hr difference between the inflection points of accumulating mRNA and increasing acetylcholinesterase activity noted previously (Meedel and Whittaker, 1983), it seems reasonable that some type of translational regulation of enzyme expression exists. If translational control is a selective mechanism involved in tissue-specific development of acetylcholine&erase, one might expect either a uniform distribution of this mRNA throughout the embryo, or some occurrence in tissues other than muscle and mesenchyme. We examined this question by purifying RNAs from 12-hr-old progeny of blastomere sets isolated at the eight-cell stage and testing their ability to direct acetylcholinesterase synthesis in the Xenopus oocyte assay (Table 2). In four experiments, no template activity significantly above the level existing in carrier RNA was detected in RNA from either animal-half or A4.1 partial embryos. The activity present in both control RNA and ascidian RNA samples probably results from nonspecific binding of endogenous oocyte

choline&erases to Immuno-Precipitin, since immunoprecipitated extracts from uninjected oocytes had a similarly low level of enzyme activity. Consistent with this suggestion was the finding that the amount of enzyme activity in immunoprecipitated extracts from uninjected oocytes was proportional to the amount of Immuno-Precipitin used. Of the partial embryos studied, only those derived from B4.1 blastomeres showed any significant template activity for acetylcholinesterase. The amounts found ranged from 28 to 121% of the level in control embryos. In three of the four experiments the B4.1 descendants have essentially the same level of template as do normal embryos. These results offer no evidence to support the idea of differentially segregated translational control elements operating on a widely distributed acetylcholinesterase mRNA. Our mRNA distribution findings agree quite well both qualitatively and quantitatively with the assays of enzyme activity in similar partial embryos. Taken together, these two types of measurement suggest very complete segregation of muscle forming

MEEDEL

DISTRIBUTION

Lineage-Specijic

AND WHITTAKER

TABLE 2 OF ACETYLCHOLINESTERASE IN 12-hr Cima EMBRYOS

mRNA

Series RNA

source

Whole embryo Animal half-embryo A4.1 quarter-embryo B4.1 quarter-embryo Carrier RNA B4.1: whole

embryo

ratio

A

B

C

D

50.7 3.3 5.3 49.7 5.6

39.1 12.2 8.0 47.2 11.1

63.6 3.1 2.2 18.0 2.3

29.2 2.9 3.4 20.6 5.0

0.98

1.21

0.28

0.71

Note. For each series, embryos were obtained from the fertilized eggs of a single animal. RNA from 25 embryos or partial embryos reared to 12 hr of age was prepared and dissolved in 5 ~1 of water. Groups of 30-35 Xenopus oocytes were injected with RNA and incubated 20 hr, or, as in Series D, 18 hr. After incubation, 25 healthy oocytes were removed, and acetylcholinesterase was purified by immunoprecipitation and assayed for 24 hr. Each oocyte was injected with 40 nl RNA; therefore, RNA from the equivalent of 5.0 embryos or partial embryos was assayed (1.0 ~1 RNA injected + 5.0 pl RNA per sample X 25 embryos). Carrier RNA-injected oocytes each received 60 ng of either L. pi&us RNA (Series A and C) or calf liver rRNA (Series B and D), which was the same type and amount of carrier RNA injected into experimental oocytes. Results are expressed as rmole (X105) acetylcholine hydrolyzed.

elements into the B4.1 lineage. Histochemical observations (Fig. 5) imply that such elements were further segregated during subsequent cleavages. DISCUSSION

Translational regulation of specific protein synthesis is seen increasingly as an important mechanism controlling embryonic development. Various examples of selective translational control have been reported (e.g., Rosenthal et ak, 1980; Fruscoloni et aL, 1983); in one case regulation of this kind was proposed to play a role in determining cell-specific patterns of protein expression in mosaic embryos, such as those of mollusc and ascidian (Collier and McCarthy, 1981). Do similar mechanisms account for the tissue-specific localization of acetylcholinesterase in ascidians? Consistent with such an interpretation is the possibility that production of functional acetylcholinesterase mRNA and its subsequent translation are not strictly coupled events (Meedel and Whittaker, 1983). We investigated this question by measuring the quantity of acetylcholinesterase mRNA in several larval cell lineages. Only in the muscle-mesenchyme lineages, where acetylcholinesterase enzymatic activity eventually becomes localized, was this mRNA detected. Translational control mechanisms were obviously not the primary cause of

Gene

Expression

485

histospecific acetylcholinesterase expression. Instead, because normal development of enzyme activity appears to require new RNA synthesis (Whittaker, 1973; Meedel and Whittaker, 1979; Meedel, 1983), our results imply that the acetylcholinesterase gene(s) were activated only in cells of the muscle and mesenchyme lineages. Regulation of enzyme expression by post-transcriptional controls (e.g., cell-specific RNA processing mechanisms) cannot, however, be ruled out. Three features of the Xenopus oocyte mRNA translation system warrant particular mention. First, mRNA-injected oocytes synthesize proteins for several days; this ability makes them much more efficient than in vitro translation systems which typically are active for only l-2 hr. Second, because oocytes produce biologically functional proteins, mRNAs that code for enzymes can be detected by the catalytic activity of their products. Not only does this result in increased sensitivity, but measuring the activity of a particular enzyme also lends specificity to the procedure. We have added a third feature to the usual oocyte microinjection methods (e.g., Soreq et aL, 1982), precipitation with an antiserum to larval acetylcholinesterase before enzymatic assay of the products. This enabled us to distinguish the ascidian enzyme translated on injected mRNA from endogenous amphibian cholinesterases (Gunderson and Miledi, 1983). By eliminating oocyte cholinesterases the immunological procedure also further increased the assay sensitivity. With the above procedure, it was possible to measure acetylcholinesterase mRNA in microsurgically dissected embryos and thus to determine the distribution of this mRNA in the developing embryo. Because acetylcholinesterase is not an abundant protein in muscle, such methods should be of general use and applicable to studying mRNAs of nearly all enzymes for which antisera can be produced. In addition to being very sensitive, this technique also provides a functional test for the presence of a specific mRNA, i.e., measurement of template activity. The value of such a test is illustrated by the recent discovery of polyadenylated mRNA-like molecules in the cytoplasm of sea urchin and frog oocytes (Costantini et aL, 1980; Anderson et aL, 1982). Although these molecules have mRNA sequences, they contain interspersed repeat sequence elements with translation stop signals in all reading frames and appear not to be translatable (reviewed by Colman, 1983). This example clearly illustrates the necessity of using criteria other than the existence of cytoplasmic mRNA sequences to define mRNA. An unexpected finding in our work was the low level of functional acetylcholinesterase mRNA in unfertilized eggs. The amount of this mRNA was about 500 times less than that found in embryos at the middle tail-

486

DEVELOPMENTAL

BIOLOGY

formation stage when acetylcholinesterase mRNA activity was highest. This maternal template seems not to be translated continuously during early development; our previous observations show that acetylcholinesterase activity remains constant before neurulation (Meede1 and Whittaker, 1979; Meedel, 1983). On the other hand, any translation products of such a rare mRNA might have been obscured by enzyme already present in the egg. Although a function for egg acetylcholinesterase cannot be excluded, the uniform distribution of this enzyme activity at the eight-cell stage indicates that it has no histospecific role in embryogenesis. Our data on acetylcholinesterase mRNA levels during development also revealed an increasing rate of accumulation of this template during each successive hourly interval between 4 and ‘7 hr after fertilization. Thus, between 5 and 6 hr more mRNA accumulated than between 4 and 5 hr, but less accumulated than between 6 and 7 hr. Because the number of strictly muscle lineage cells also increases throughout gastrulation (Conklin, 1905), this pattern could result from more cells producing acetylcholinesterase mRNA. Such an explanation is consistent with the expectation that differentiated characteristics become expressed only after cell lineages have completely sorted out, thereby preventing phenotype-specific features from occurring in extraneous tissues. Assays of enzymatic activity and functional mRNA both indicated that acetylcholinesterase expression was confined to partial embryos derived from cells of the B4.1 lineage. These results are entirely in agreement with a strict segregation of muscle determinants into the B4.1 blastomere pair, and although diffusible gradient models of determinant localization (Slack, 1983) cannot be excluded, they are not required to explain our findings. Investigations of the ascidian cytoskeleton suggest that cytoskeletal anchorage mechanisms might account for determinant localizations and segregations (Jeffery and Meier, 1983). Such anchorages would be capable of producing the sharp boundaries of developmental fate that seem to occur in ascidian development (Ortolani, 1954). Results of a recent study of ascidian cell lineages using microinjected horseradish peroxidase suggested that some muscle cells originate from b4.2 and A4.1 blastomere pairs (Nishida and Satoh, 1983). According to these authors, b4.2 and A4.1 descendants comprise 20-25s of the muscle cells in Ciona embryos. Conceivably, contribution from these “extra” cells could account for the slightly lower level of enzyme activity sometimes observed in B4.1-derived partial embryos compared to whole embryos (Table 1). Considering the restricted number of embryos (5-10) used in our assays, we cannot on the basis of these measurements alone rule

VOLUME

105, 1984

out some contribution to muscle from cells other than those of the B4.1 lineage. More difficult to understand is the complete absence of any form of acetylcholinesterase expression in partial embryos resulting from animal-half and A4.1 blastomeres which would be predicted from the experiments of Nishida and Satoh (1983). Some of this difference may be related to their use of whole embryos and our use of partial embryos-a view consistent with some of our work in progress (J. R. Whittaker, R. J. Crowther, and T. H. Meedel, 1984, in preparation). We have observed that three-quarter partial embryos, composed of animal-half and A4.1 lineages (but missing the B4.1 descendants), develop a much higher level of enzyme activity than is found when the activity in A4.1 and animal-half partial embryos raised separately are added together. Some muscle cells may indeed arise from blastomeres other than the B4.1 pair, but phenotypic differentiation in cells from these secondary muscle sources appears to depend on inductive interactions; developmental autonomy of muscle expression is not one of their features. Differentiation of muscle in cells from these other lineages appears not to involve the same cytoplasmic determinant mechanisms that control expression in the lineages derived from the B4.1 blastomeres. Our present results confirm and extend earlier findings concerned with the autonomy (self-differentiation) of muscle development in partial embryos containing the muscle cell lineages. By assaying the functional mRNA for a muscle gene we now have a more direct measure of gene expression by these cells than the occurrence of acetylcholinesterase activity used previously (Whittaker et aL, 1977; Whittaker, 1982). In addition to acetylcholinesterase, myofilaments and myofibrils characteristic of muscle are also expressed when blastomeres of this lineage are raised in isolation (Crowther and Whittaker, 1983). Presumably, such self-differentiation is caused by the segregation of egg cytoplasmic determinants into the muscle lineage blastomeres where they exert a strong influence on subsequent development. By refinement and extension of current work on methods for assaying myofibrillar proteins (Meedel, 1983), it should soon be possible to analyze quantitatively the expression of myofibrillar proteins in isolated partial embryos. Together with studies of acetylcholinesterase such experiments should permit further insights into the nature and complexity of the system controlling ascidian muscle development. This investigation was supported by Grant HD-16547 from National Institute of Child Health and Human Development, March of Dimes Birth Defects Foundation Grant l-780.

the and

MEEDEL

AND WHITTAKER

Linewe-Spec$ie

REFERENCES ANDERSON, D. M., RICHTER, J. D., CHAMBERLIN, M. E., PRICE, D. H., BRITTEN, R. J., SMITH, L. D., and DAVIDSON, E. H. (1982). Sequence organization of the poly(A)RNA synthesized and accumulated in lampbrush chromosome stage Xenqpus lo&s oocytes. J. MoL BioL 155, 281-309. COLLIER, J. R., and MCCARTHY, M. E. (1981). Regulation of polypeptide synthesis during early embryogenesis of Ilynassa obsoleta. D@erentiation 19, 31-46. COLMAN, A. (1983). Maternal mysteries. Trends Biochem Sci. 8, 37. CONKLIN, E. G. (1905). The organization and lineage of the ascidian egg. J Acad NatL Sci. (Philadelphia) 13, 1-119. COSTANTINI, F. D., BRITTEN, R. J., and DAVIDSON, E. H. (1980). Message sequences and short repetitive sequences are interspersed in sea urchin egg poly(A+)RNAs. Nature (London) 287, 111-117. CROWTHER, R. J., and WHITTAKER, J. R. (1983). Developmental autonomy of muscle fine structure in muscle lineage cells of ascidian embryos. Dev. BioL 96, l-10. DAVIDSON, E. H. (1976). “Gene Activity in Early Development,” 2nd ed. Academic Press, New York. DUMONT, J. N. (1972). Oogenesis in Xenqpus Levis (Daudin). 1. Stages of oocyte development in laboratory maintained animals. J. Morph& 136, 153-179. DURANTE, M. (1956). Cholinesterase in the development of Ciona intestinalis (Ascidia). Experientia 12, 307-308. FRUSCOLONI, P., AL-ATIA, G. R., and JACOBS-LORENA, M. (1983). Translational regulation of a specific gene during oogenesis and embryogenesis of Drosophila. Proc. NatL Acad Sci. USA 80, 33593363. GUNDERSON, C. B., and MILEDI, R. (1983). Acetylcholinesterase activity of Xenow laevis oocytes. Neuroscience 10,1487-1495. GURDON, J. B. (1977). Methods for nuclear transplantation in amphibia. In “Methods in Cell Biology” (G. Stein, J. Stein, and L. J. Kleinsmith, eds.), Vol. 16, pp. 125-139. Academic Press, New York. GURDON, J. B. (1981). Concepts of gene control in development. In “Developmental Biology Using Purified Genes” (D. D. Brown, ed.), pp. l-10. Academic Press, New York. JEFFERY, W. R., and MEIER, S. (1983). A yellow crescent cytoskeletal domain in ascidian eggs and its role in early development. Dev. BioL 96, 125-143. KARNOVSKY, M. J. (1965). A formaldehyde-glutaraldehyde fixative of high osmolality for use in electron microscopy. J Cell BioL 27, 137A-138A. KARNOVSKY, M. J., and ROOTS, L. (1964). A “direct-coloring” thiocholine method for cholinesterase. J. Histochem Cytochem. 12, 219221. KESSLER, S. W. (1975). Rapid isolation of antigens from cells with a staphylococcal protein A-antibody adsorbent: Parameters of the interaction of antibody-antigen complexes with protein A. J. Immunol 115, 1617-1624. LEE, D. C., MCKNIGHT, G. S., and PALMITER, R. D. (1978). The action of estrogen and progesterone on the expression of the transferrin gene. J BioL Chem. 253,3494-3503. MEEDEL, T. H. (1980). Purification and characterization of an ascidian larval acetylcholinesterase. Biochim Biuphys. Acta 615, 360-369. MEEDEL, T. H. (1983). Myosin expression in the developing ascidian embryo. J. Exp. Zoo1 227, 203-211. MEEDEL, T. H., and WHITTAKER, J. R. (1978). Messenger RNA

Gene

Expression

487

synthesis during early ascidian development. Dev. BioL 66, 410421. MEEDEL, T. H., and WHITTAKER, J. R. (1979). Development of acetylcholinesterase during embryogenesis of the ascidian Ciana intestinalis. J. Exp. ZooL 210, l-10. MEEDEL, T. H., and WHITTAKER, J. R. (1983). Development of translationally active mRNA for larval muscle acetylcholinesterase during ascidian embryogenesis. Proc. NatL Acad Sci USA 80, 4761-4765. NISHIDA, H., and SATOH, N. (1983). Cell lineage analysis in ascidian embryos by intracellular injection of a tracer enzyme. 1. Up to the eight-cell stage. De-v. BioL 99, 382-394. ORTOLANI, G. (1954). Risultati definitivi sulla distribuzione dei territori presuntivi degli organi de1 germe di Ascidie allo stadio VIII, determinati con le marche al carbone. PubbL Staz. ZooL Napoli 25, 161-187. ORTOLANI, G. (1955). The presumptive territory of the mesoderm in ascidian germ. Experientia 11, 445-446. PELHAM, H. R. B., and JACKSON, R. J. (1976). An efficient mRNAdependent translation system from reticulocyte lysates. Eur. J B&hem. 67, 247-256. PERRY, H. E., and MELTON, D. A. (1983). A rapid increase in acetylcholinesterase mRNA during ascidian embryogenesis as demonstrated by microinjection into Xenopus luevis oocytes. Cell D$er. 13, 233-238. REVERBERI, G. (1971). Ascidians. In “Experimental Embryology of Marine and Freshwater Invertebrates” (G. Reverberi, ed.), pp. 507-550. Elsevier/North-Holland, New York. ROSENTHAL, E. T., HUNT, T., and RUDERMAN, J. V. (1980). Selective translation of mRNA controls the pattern of protein synthesis during early development of the surf clam Spisula solidissima Cell 20, 487-494. SCHRIER, B. K., WILSON, S. H., and NIRENBERG, M. (1974). Cultured cell systems and methods for neurobiology. In “Methods in Enzymology” (S. Fleischer and L. Packer, eds.), Vol. 32, Part B, pp. 774-777. Academic Press, New York. SLACK, J. M. W. (1983). “From Egg to Embryo.” Cambridge Univ. Press, Cambridge. SOREQ, H., PARVARI, R., and SILMAN, I. (1982). Biosynthesis and secretion of catalytically active acetylcholinesterase in Xenqpus oocytes microinjected with mRNA from rat brain and from Torpedo electric organ. Proc. NatL Acad Sci. USA 79, 830-834. TOMMERUP, N. (1982). Precision-bored microcapillaries for microinjection of nano-litre quantities without precalibration. Exp. Cell Res. 140, 476-479. WHITTAKER, J. R. (1973). Segregation during ascidian embryogenesis of egg cytoplasmic information for tissue-specific enzyme development. Proc. NatL Acad. Sci. USA 70, 2096-2100. WHITTAKER, J. R. (1979). Cytoplasmic determinants of tissue differentiation in the ascidian egg. In “Determinants of Spatial Organization” (S. Subtelny and I. R. Konigsberg, eds.), pp. 29-51. Academic Press, New York. WHITTAKER, J. R. (1980). Acetylcholinesterase development in extra cells caused by changing the distribution of myoplasm in ascidian embryos. J. EmlnyoL Exp. MorphoL 65,343-354. WHITTAKER, J. R. (1982). Muscle lineage cytoplasm can change the developmental expression in epidermal lineage cells of ascidian embryos. Dev. BioL 93,463-470. WHITTAKER, J. R., ORTOLANI, G., and FARINELLA-FERRUZZA, N. (1977). Autonomy of acetylcholinesterase differentiation in muscle lineage cells of ascidian embryos. Dev. BioL 55, 196-200.