Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
AMER. ZOOl... , 35:358-371 (1995) Target Recognition by Mesenchyme Cells During Sea Urchin Gastrulation l JEFF HARDIN Department ofZoology, University of Wisconsin, 1117 W. lohnson St., Madison, Wisconsin 53706 SYNOPSIS. During sea urchin gastrulation, mesenchyme cells must migrate to the correct locations and attach to specific sites within the embryo. Primary mesenchyme cells (PMCs) aggregate into two major clusters in the ventrolateral region of the embryo and other PMCs form an elaborate pattern that presages the skeletal pattern of the larva. Numerous exper iments, including alteration of dorsoventral differentiation of the ecto derm with NiCl z, indicate that the ectoderm provides much of the pattern information to which these cells respond. Secondary mesenchyme cells (SMCs) attach to the ectoderm at a specific site near the animal pole at the end ofgastrulation, undergoing a dramatic change in protrusive behav ior as they do so. Experiments indicate that the target region to which they respond is restricted to a small domain that will form the larval mouth, and that this region is unique in its ability to elicit these changes in behavior. A third population of patterned mesenchyme has recently been identified by their expression of the sea urchin homologue of the Drosophila transcription factor, snail. Snail expressing cells form two bilateral clusters near the sites where the larval arms will form as gastru lation ends. Finally, we have identified two interactions between cells in the early embryo that seem to be important for the establishment of pattern information for mesenchyme: (1) interactions between macromere descendants and neighboring cells can establish new bilateral patterning sites for PMCs, and (2) interactions betw~en the animal pole and cells in a more vegetal location appear to result in induction of the oral field and associated pattern information for both PMCs and SMCs. in reliable locations. One of the great out standing questions in the study of gastru lation is how mesenchymal cells within the gastrula engage in such coordinated mor phogenetic movements as the incipient body plan emerges, and how the response of such cells to pattern information in the embryo modulates these behaviors. Several model systems have been used to address this question. Well studied examples of the con trol of mesenchymal migration during gas trulation include the deep cells that form the embryo proper in teleosts (Wourms et al., 1972; Trinkaus et aI., 1991; Ho, 1992), the ingression ofcells from the presumptive epiblast to form the posterior margin of the developing avian embryo (Stern, 1990, From the Symposium The Role ofCell-Cell lnter 1991) and the migration of head mesoderm actions and Environmental Stimuli in the Development ofMarine Invertebrates presented at the Annual Meet cells in the amphibian embryo (Winklbauer et al., 1991; Keller and Winklbauer, 1992; ing of the American Society of Zoologists, 27-30 December 1993. Los Angeles, California. Boucaut et al., 1991). Another model sys 358 INTRODUCTION The movements ofgastrulation are a par ticularly dramatic example of the response of tissue sheets and freely migrating mes enchymal cells to pattern information within the embryo. As tissue sheets involute, invaginate or spread during gastrulation, they do so based on precise partitioning of tissue territories in which different morpho genetic programs are carried out. The same is true of mesenchymal cells; different pop ulations of mesenchymal cells respond very differently to pattern information within the embryo, resulting in the migration and aggregation ofvarious types ofmesenchyme I TARGET RECOGNITION BY MESENCHYME DURlNG SEA URCHIN GASTRULATION 359 ventral +---"""'i~~ dorsal (aboral) (oral) animal secondary mesenchyme apical plate archenteron vegetal view from oral side spicule rudiment lateral view late gastrula/prism FIG. I. A schematic representation of pallerning sites for mesenchyme in the sea urchin late gastrula, viewed from the the ventral side (left) and the left side (right). For description of labeled structures, see the tex\. invagination), the archenteron extends lf4 '12 of the way across the blastocoel. A short pause follows primary invagination, after which the archenteron resumes its elonga tion (secondary invagination). At about the time secondary invagination begins, cells at the tip of the archenteron (secondary mes enchyme cells) become protrusive, extend ing long filopodia into the blastocoel. Even tually the archenteron elongates across the blastocoel, and its tip attaches to the ventral ectoderm near the animal pole (Fig. 1). As the archenteron invaginates, primary mes enchyme cells migrate and eventually local ize in a precise pattern. Two major bilateral clusters ofPMCs appear in the ventrolateral region of the embryo, near two bilateral thickenings in the ectoderm, so that the embryo has a morphologically obvious dor soventral axis (Fig. 1). Other PMCs form a OVERVIEW OF GASTRULATION IN THE ring in the vegetal half of the embryo, which SEA URCHIN EMBRYO connects with the ventrolateral clusters in a Just prior to gastrulation in the sea urchin characteristic fashion (see Ettensohn, 1990 embryo, the epithelium at the vegetal pole for an in-depth discussion of skeletal pat of the blastula flattens and thickens to form terning). The PMCs eventually fuse with one the vegetal plate. Following the flattening of another to form a syncytium, and they the vegetal plate, primary mesenchyme cells secrete the skeletal rods, or spicules, of the ingress into the blastocoel. Then the vegetal larva (for reviews of skeleton formation in plate begins to bend inward to form a short, the sea urchin larva see Decker and Len squat cylinder, the archenteron. During this narz, 1988; Benson and Wilt, 1992; Etten initial phase of invagination (primary sohn and Ingersoll, 1992). lntimately the tern in which the patterning and morpho genesis of mesenchyme has been studied is the sea urchin embryo. Because of its simple organization, the ability to observe mor phogenetic movements directly in living embryos, and because of its rich historical context, the sea urchin embryo is a good system in which to study ways in which axial patterning influences the behavior of mesenchyme at the single-cell level (for reviews, see Wilt, 1987; Ettensohn, 1991; Hardin, 1990, 1994a, b). In the following sections, classical and new experimental evidence for regionally specific patterning of the ectoderm of the sea urchin embryo will be presented. In addition, experiments that indicate that cell interactions are required for the emergence of this pattern information will be discussed. 360 JEFF HARDIN tip of the archenteron fuses with the ventral ectoderm near the animal pole to form the larval mouth. As the pluteus larva develops, the archenteron becomes tripartite, and undergoes regional differentiation to form an esophagus, stomach, and intestine. EcrODERMAL CuEs FOR LoCALIZATION OF PRIMARY MESENCHYME CELLS After their ingression into the blastocoel, primary mesenchyme cells (PMCs) remain at the vegetal plate briefly before migrating within the interior of the embryo. The basic mechanisms of migration of these cells have been extensively reviewed elsewhere (Gus tafson and Wolpert, 1963, 1967; Solursh, 1986; Ettensohn and Ingersoll, 1992; Har din, 1990, 1994a, b). It is clear from a num ber of studies that PMCs respond to pattern information present within the ectoderm during gastrulation. The precise ring pattern formed by these cells, the aggregation of PMCs into two ventrolateral clusters, and the elaboration of details of skeletal pattern are all influenced by ectodermal cues. One of the earliest cues for PMC localization appears to reside in the vegetal plate. Fol lowing their ingression, PMCs remain at the vegetal pole of the embryo for a noticeable period of time prior to their migration away from the vegetal plate. If PMCs are dis placed to the animal pole by centrifugation (Okazaki et al., 1962) or transplanted into the animal pole (Ettensohn and McClay, 1986), or when PMCs arise from single micromeres incorporated ectopically (Wray and McClay, 1988), they migrate to the veg etal plate region and remain there until undisturbed PMCs migrate away from the vegetal plate (Ettensohn and McClay, 1986). Subsequent aspects ofPMC patterning also appear to be conditioned by interaction of PMCs with the ectoderm. Okazaki and col leagues noticed that the ventrolateral clus ters form atop or very near regions of thick ened ectoderm that produce an optical effect reminiscent of a Japanese fan (Okazaki et aI., 1962). When these thickenings are dis placed by vegetalizing with lithium chlo ride, the ring and clusters also shift (Wolpert and Gustafson [1961]; see the article by Wilt et al. [1995] for further information on the effects of LiCI on sea urchin develop ment). Gustafson and Wolpert (1963 , 1967) suggested that the interaction ofPMCs with the ventrolateral thickenings was largely due to physical considerations, where presumed adhesive sites would be found at a higher density. Other distributed guidance cues for PMCs within the em bryo may exist, based on stud ies of pattern regulation. The classic studies of Driesch (1900) and careful studies by Takahashi et al. (1979) indicate that in 112 and 1/4 dwarf larvae, the skeleton is capable of size regulation. Even more dramatic evi dence for such regulative control was pro vided by Ettensohn. When 2-3 times the normal number of PMCs are transplanted into a host embryo, the resulting skeletal pattern is nonetheless normal (Ettensohn, 1990). To provide a direct test of the extent to which the ectoderm influences patterning of PMCs, we have used radialized embryos as a tool for studying patterning of primary mesenchyme cells. In particular, we have used nickel chloride treatment to produce embryos in which all traces of bilateral sym metry are abolished (Hardin et aI., 1992). Based on in situ hybridization and immu nostaining, one of the primary effects of nickel seems to be to alter the distribution of ectodermal tissues in the embryo, result ing in an overproduction of ventral (oral) ectoderm and a corresponding decrease in the amoun t of dorsal (aboral) ectoderm that is produced (Hardin et aI. , 1992). The cil iated band, which normally forms at the boundary between dorsal and ventral ecto derm, is concomitantly shifted to a vegetal position. Embryos radialized in this fashion using nickel chloride are a useful tool for studying ectodermal cues for PMC pattern ing for several reasons. Since the ventrolat eral clusters normally form as two foci near the vegetal pole close to the position of the ciliated band, the position of spiculogenic clusters should be shifted in nickel treated embryos concomitant with the new posi tions ofectodermal territories in the embryo, if these anatomical markers are reliable pre dictors of sites of pattern formation. In addition, it is possible to perform reciprocal transplants ofPMCs between nickel-treated and normal embryos to test whether or not TARGET RECOGNITION BY MESENCHYME DURING SEA URCHIN GASTRULATION 361 A .. ~ ; ,J , , 4 >:~' .. " "\0. .-. \ ~ FIG, 2. Effects of Ni0 2 on skeletal pattern in Lytechinus variegatus visualized with polarization microscopy (adapted from Hardin et al,. 1992, with permission) (a) Prism stage embryo radialized with 0.5 mM NiCI 2 • (b) Normal pluteus. Bar = 20 /Lm. the ectoderm is a primary source of pattern information for spiculogenesis. Skeletal pattern is severely disrupted in nickel-treated embryos. Rather than form ing the normal two bilateral triradiate skel etal elements, nickel-treated embryos pro duce as many as a dozen small triradiate elements, disposed in a radial pattern near the vegetal margin of the embryo (Fig. 2). As these skeletal elements grow, they result in an embryo with severe disruption ofskel etal pattern, although the total number of PMCs and the amount of skeletal material produced appears to be normal (Hardin et al.) 1992). To determine whether the pri mary defect in nickel-treated embryos resides with their PMCs or with the ecto dermal environment, reciprocal transplants were performed between nickel-treated and normal embryos (Armstrong et al., 1993). When a normal complement ofPMCs from nickel-treated embryos is transplanted into normal host embryos from which all PMCs have been previously removed, the result ing skeletal pattern is completely normal. Conversely, when PMCs are removed from nickel-treated embryos and a full comple ment of normal PMCs is subsequently added, the skeletal pattern is completely radialized. These results indicate that the source of the ectoderm largely determines the pattern that the PMCs adopt, at least regarding bilaterality. In addition, by repeating the experiments of Ettensohn (1990) using nickel-treated rather than nor mal embryos, it can be shown that skeletal size and pattern are still regulated in radi alized embryos, despite obvious disruptions in skeletal morphology (Armstrong et al., 1993). As discussed below, other regionally specific aspects of skeletal pattern are also under ectodermal control. However, finer aspects of skeletal pattern, e.g., the intricate fenestrations that are elaborated along the length of various skeletal rods, appear to be an endogenous characteristic of the species from which the PMCs are derived, based on heterospecific transplants (Armstrong and McClay, 1994). EcrODERMAL CuEs FOR LocALIZATION OF SECONDARY MESENCHYME CELLS Time-lapse cinemicrographic studies by Gustafson and coworkers thoroughly doc umented the basic protrusive behavior of SMCs in Psammechinus miliaris (reviewed by Gustafson and Wolpert, 1963, 1967). These studies suggested that the filopodia of secondary mesenchyme cells "ran domly" explore the blastocoel, undergoing continual cycles of extension, attempted attachment, and retraction. As gastrulation 362 JEFF HARDIN ends SMCs undergo a change in their behav ior; much of their protrusive activity sub sides, and they may become more loosely associated with the tip of the archenteron (Dan and Inaba, 1968; Gustafson and Kin nander, 1960). Eventually, several hours after gastrulation is over, the tip of the archenteron attaches to the stomodeum (an invagination of the oral ectoderm) and fuses with it to form the mouth of the pluteus larva (Gustafson and Kinnander, 1960). Attachments ofSMCs during gastrulation Given the apparently random explora tory behavior ofSMCs, how do they reliably attach the tip of the archenteron to the ani mal pole region as gastrulation ends? Classic observations raised several possibilities. Gustafson suggested physical cues were largely responsible, including (l) differences in surface topography at contact sites between epithelial cells (Gustafson, 1963), (2) tissue proximity (Gustafson and Wol pert, 1967; Gustafson, 1969), and (3) tissue curvature, which might allow access of fil opodia to non-specific adhesive sites in some regions but not others (Gustafson, 1963; reviewed by Gustafson and Wolpert, 1967). In addition, Gustafson suggested that the ventral ectoderm might be generally more adhesive than the dorsal ectoderm, which would result in localization of both primary mesenchyme cells and later, secondary mes enchyme cells (Gustafson, 1969). Other observations, however, suggested that there might be more specific guidance cues resi dent near the animal pole. In several Jap anese species (Dan and Inaba, 1968) and in Lytechinus variegatus (reviewed in Trin kaus, 1984) filopodia initially extend lat erally, and only late in gastrulation do they extend upward. This suggested that although the essential motile program of SMCs involves continual cycles of random filo podial extension, attachment and eventual withdrawal, there might be unique pattern information to which they respond near the animal pole. A n animal pole "target" exists for SMCs We reexamined the attachment ofSMCs to the animal pole region in some detail (Hardin and McClay, 1990). At the end of gastrulation the tip of the archenteron makes contact with the ectoderm near a thickened region of epithelium, the apical plate (cf Fig. 1). At this time the exploratory behav ior of the filopodia largely ceases, marking the end of gastrulation. Based on real-time analysis of the residence times of attached filopodia, protrusions that make contact with the ectoderm in the apical plate region remain attached 20-50 times longer than attachments observed at any other site along the blastocoel wall (Hardin and McClay, 1990). The SMCs bearing the long-lived fil opodia eventually change their behavior as they flatten and spread onto this region. In some species, such as Lytechinus variegatus, this region lies near the animal pole; in other species, such as Strongylocentrotus purpur atus, it is located on the ventral side of the animal hemisphere (Hardin and McClay, 1990). The normal behavior of SMCs suggests that the animal pole region serves as a "tar get" for filopodial attachment, and several experiments indicate that SMCs do in fact respond uniquely to this region. First, when the animal pole region is pushed toward the tip of the archenteron with a micropipette (Fig. 3) or when embryos are trapped in Nitex mesh of the appropriate dimensions, SMCs interact with the animal pole pre cociously, and their exploratory behavior ceases ahead of schedule. In some cases, SMCs make a precocious, stable attachment to the animal pole (Fig. 3; also see Hardin and McClay, 1990). SMCs make transient contacts with other areas of the blastocoel wall when they are indented, but the archen teron continues past such indentations, eventually attaching to the usual site (Har din and McClay, 1990). Second, when con tact of SMCs with the animal pole is pre vented by extruding embryos into capillary tubing so that filopodia extended by SMCs cannot reach the animal pole, SMCs con tinue the cyclical extension of filopodia for abnormally long periods of time (Hardin and McClay, 1990). If the archenteron is prevented from reaching the animal pole for several hours, some SMCs detach from the archenteron, migrate to the animal pole, and TARGET RECOGNITION BY MESENCHYME DURING SEA URCHIN GASTRULATION 363 Release Elongate / No release Dent at uteral dent Q=-.-- at FIG. 3. Experiments by Hardin and McClay (1990) demonstrating the presence of an animal pole "target" for secondary mesenchyme cells (adapted from McClay et al., 1992, with permission). Indentation of the animal pole with a micropipette results in precocious attachment of the tip of the archenteron to the animal pole region; elongation of an embryo in capillary tubing prevents attachment of the archenteron to the animal pole. In elongated embryos, some SMCs migrate as individuals to the animal pole, where they aggregate (arrow). Bars = 20 /-Lm. undergo the change in behavior seen in nor mal embryos (Fig. 3; Hardin and McClay, 1990). Thus, SMCs appear to be pro grammed to continue extending filopodia until the appropriate target is reached on the wall of the blastocoel. The animal pole region of the embryo focuses filopodial attachments as gastrula tion nears its conclusion. Since filopodia can attach to the animal pole several hours before they typically do and for several hours after they normally do, target recognition allows for temporal flexibility during gas trulation. Target recognition by SMCs has other consequences as well: the animal pole region and archenteron bend towards the ventral (oral) side of the embryo after gas trulation (Gustafson and Kinnander, 1960; Dan and Inaba, 1968), thereby moving the archenteron close to the site where it fuses with the stomodeum . Spatial extent of the target Based on time-lapse cinemicrography, the region to which SMCs respond is fairly small, on the order of 20 ~m (Hardin and McClay, 1990), and appears to correspond well with the region later occupied by the oral attach ment that generates the mouth. More pre cise mapping can be performed to establish the spatial extent of the target region by transplanting rhodamine-labeled SMCs into unlabeled hosts. When cells are removed from the tip of the archenteron of a labeled donor embryo and transplanted into a host midgastrula, the transplanted cells migrate and take part in apparently normal pattern forming events. Discrete clusters of the 364 JEFF HARDIN FIG. 4. Spatial extent of the animal pole target revealed by transplants of rhomdamine-Iabeled SMCs into unlabeled host embryos. (a) bright field, (b) epifiuorescence. RITC-Iabeled cells aggregate in a focal spot near the animal pole (arrow). Bar = 20 I'm. transplanted cells form a rosette pattern near the animal pole, in precisely the location where the archenteron attaches at the end of gastrulation (Fig. 4). This region is quite smaIl, on the order of 15-20 jlm, agreeing well with the previous estimates based on differences in filopodial behavior. Phylogenetic variations in target placement and utilization during gastrulation It seems clear that the shape of the embryo at the onset of gastrulation can impose sig nificant constraints on how gastrulation progresses. In the normal Lytechinus var iegatus embryo, successful filopodial at tachments are restricted to lateral regions of the embryo early in gastrulation (Trinkaus, 1984; Hardin and McClay, 1990). Numer ous experiments indicate that this is due to the shape of the embryo; the animal pole is simply too far from the tip of the archen teron for SMCs to make successful attach ments to it (Hardin and McClay, 1990). However, a second morphogenetic process, autonomous extension of the archenteron, operates concurrently with the onset of fil opodial exploration (Hardin, 1988, 1989). As the archenteron extends, the probability of a filopodium/target encounter is greatly enhanced by the proximity of the animal pole to the tip of the archenteron at the 2/3 % gastrula stage. As stable filopodia remain attached they puB the archenteron even closer towards the animal pole, thereby giv ing more filopodia opportunities to make stable contacts with the animal pole region. By combining a small set of relatively sim ple ceIl behaviors (i.e., autonomous cell rearrangement, random exploration by fil opodia, and target recognition), sea urchin gastrulation is a "robust" process that can be successfully completed despite wide vari ations in embryonic shape and the positions of interacting tissues. A survey of gastrulation in a number of sea urchin species reinforces this notion. Dif ferent species rely on varying combinations of these processes during archenteron elon gation, and these differences appear to arise in part from differences in embryonic shape (Fig. 5; Hardin and McClay, 1990). Modes of archenteron elongation include "central elongation," in which the archenteron is equidistant from all lateral ectodermal sur faces (e.g., Lytechinus variegatus, L. pictus), "dorsal crawling," in which the dorsal ecto derm is near the tip of the archenteron (Psammechinus miliaris, Echinus microtu berculatus), and "ventral crawling," in which the ventral side is closer (e.g., S. purpura IUS). In still other species, such as the cida roid, Eucidaris tribuloides, and the Japanese sand dollar, Clypeaster japonicus, filopodia may extend laterally towards the stomodeal invagination directly, in a manner similar TARGET RECOGNITION BY MESENCHYME DURING SEA URCHIN GASTRULATION v D v D Ventral crawlers ( Strongylocentrotus) Central elongation ( Lytechinus ) v 365 D Dorsal crawlers (Echinus, Psammechinus ) V? D? Central elongation ( Eucidaris , M espilia ) FIG. 5. Phylogenetic differences in archenteron elongation and oral contact among various species ofeuechinoids and cidaroids (from Hardin and McClay, 1990, with permission). V = ventral, D = dorsal. to mouth formation in asteroids. In all of ila (Boulay el al., 1987; Alberga el aI., 1991; these cases, the simple cell behaviors that Leptin, 1991; Ip el al., 1992; Whiteley el are responsible for archenteron elongation . al., 1992), Xenopus (Sargent and Bennett, and attachment appear to be flexible enough 1990), zebrafish (Hammerschmidt and to permit such phylogenetic variations in Niisslein-Volhard, 1993; Thisse et aI., 1993), embryonic shape. chick (Nieto el al., 1994) and mouse (Nieto el al., 1992; Smith el aI., 1992). Based on A SEA URCHIN HOMOLOGUE OF THE sequence analysis, the only highly con DROSOPHILA PROTEIN SNAIL Is served region of the molecule in all of these EXPRESSED BY MESENCHYME cases is the putative DNA binding domain, Recently, we have identified another pop which contains five zinc finger loops in all ulation of patterned mesenchyme by of the species examined. In zebrafish there searching for sea urchin orthologues of appear to be two snail-related proteins, one mesodermal transcription factors expressed of which lacks conservation ofthe first loop in phylogenetically diverse organisms. One (Thisse el al., 1993); in mice, the protein such transcription factor is snail. Members characterized so far only possess four loops of this family of zinc finger proteins have (Nieto el al., 1992; Smith el al., 1992). We been characterized extensively in Drosoph have used the DNA binding domain of the 366 JEFF HARDIN Drosophila snail orthologue to clone and characterize a sea urchin member of this gene family (Fig. 6a) Based on whole mount in situ hybridization, sea urchin snail (usna) transcripts are first detectable midway through gastrulation in the archenteron. By the end of gastrulation, a cluster of cells in the tip of the archenteron, disposed asym metrically with respect to the right-left axis of the embryo, express usna transcripts. After gastrulation, two clusters of 6-10 mes enchyme cells express usna transcripts near the sites where the arm buds will grow shortly after gastrulation (Fig. 6b). By the prism stage, staining disappears at the tip ofthe archenteron, and intense staining per sists in the two clusters, but eventually dis appears by the pluteus stage. It is not clear whether cells expressing usna are a subset of PMCs or a subpopulation of SMCs that have left the vegetal plate or the archenteron during gastrulation. Interestingly, staining of what appears to be the same clusters is obtained for the myogenic factor SUM -1 (J. Venuti, personal communication). What function these clusters ofcells play, whether or not they arise by migration of cells away from the tip of the archenteron, and how their bilateral pattern is produced are unknown at present. Wilt, 1990; Davidson, 1989). It is also clear that local cell-cell interactions between cells from different tissue territories can influ ence the expression of particular cell fates in dramatic ways. One well-studied example of cell inter actions during early sea urchin developmen t is the influence of micro meres on the devel opment of cells with which they are in con tact. Two potential effects of micro meres on nearby cells have been documented. First, by studying the development of isolated mesomeres, Wilt and colleagues have shown that mesomeres can respond to contact with micromeres by producing gut and skeletal structures (see the article in this volume by Wilt et al., 1995). Second, transplants using 16- and 32-cell embryos by Horstadius (1935) and Ransick and Davidson (1993) have shown that micromeres in contact with meso meres can induce presumptive ecto derm to form an archenteron, even though they do not normally do so. In both sorts of experiments, it has also been shown that the induced tissues express the appropriate marker mRNAs and/or proteins (Khaner and Wilt, 1990; Ransick and Davidson, 1993; see the article by Wilt et al., 1995). In the case of micromere transplants into the animal poles of recipient embryos, both Horstadius (1935) and Ransick and David son (1993) have shown that an additional CELL INTERACTIONS REGULATE THE consequence of the implantation of the ApPEARANCE OF ECTODERMAL ectopic micromeres is the production ofone PATIERN ELEMENTS or more supernumerary skeletal elements. Macromere descendants can induce new In the carefully controlled studies of Ran patterning sites for PMCs sick and Davidson (1993), a complete, bilat The pattern information associated with eral skeleton can form straddling the ectop the ectoderm during gastrulation in the sea ically induced archenteron. The results of urchin embryo depends on events during Ransick and Davidson suggest several pos early development. Studies by Horstadius sible mechanisms by which these appar (1935; reviewed in Horstadius, 1939, 1973) ently novel patterning sites arise: (1) they and Cameron and colleagues indicate that could arise via inductive signals produced the major tissue territories of the embryo by the micromeres; (2) they could arise via arise from founder cells that are born by the sequential inductions, in which presump 5th-6th cleavage (reviewed in Cameron and tive gut tissue is first induced, and then the Davidson, 1991). Equally clear is the famous induced tissues sends lateral signals to induce ability of sea urchin embryos to exhibit reg new patterning sites, or (3) a combination ulative development when blastomeres from of these mechanisms could be required. Two pieces of experimental evidence the early embryo are isolated and recom bined (see the article by Wilt et al. in this indicate that the second of these possibili volume; 1995 also see reviews by Horstad ties is at least sufficient to account for the ius, 1939, 1973; Wilt, 1987; Livingston and induction ofpatterning sites within the ecto TARGET RECOGNITION BY MESENCHYME DURING SEA URCHIN GASTRULATION 367 A • s na J 1 u sna x s na KN'IRFKCOECQKHYSTSMGLSKHRQFHCPA.' .i'CNQEKKTHSCEECGKLYTTIGALKMHIR 300 DS TKYHC POCGKEYSTFGGLSKHRQLHCDA- - - -QNKKTFNCKYCDKEYMSLGALKMHIR 2 46 EA EKFQCNL CSKSYSTFAGLSKH.KQLHCDS- - - -QTRKSFSCKYCEKEYVSLGALKMHIR 1) 0 sna i I THTLPCKCPICGKAFSRPWLLQGHIRTHTGEKPFQCPOCPRSFADRSNLRAHQQTHVDVK 36 0 us n a THTLPCKCKFCGKAFSRPWLLQGHIRTHTGEKPFSCPHCQRAFADRSNLRAHLQTHSEVK )0 6 xsn a SHT~.PCYCKTCGKAFSRPWL'L QGHIRTHTGEKPFSCTHCNRAFADRSNLRAHLQTHSDVK 230 390 snall KYACQVCHKSFSRMSLLNKJi SSSNCTI TI A 33? us n a KYSCKSCGKTFSRMSLLNKHEESGCISSGSD 25 9 xs na KYQCKSCSRTFSRMSLLHKHEETGC - - TVAH FIG. 6. Characterization of a population of cells expressing transcripts of a sea urchin member of the snail family of proteins. (a) sequence comparison within the zinc-finger domain ofconceptual translations of Drosophila snail, Xenopus laevis snail (xsna), and Lylechinus variegalus snail (usna). Identical amino acids are shaded. The arrow marks the start of the first of the five canonical zinc finger loops. (b) Whole mount in silu hybridization of a L. variegalus prism stage embryo. Intense staining of two clusters of mesenchyme is evident (arrows). Bar = 20 !lm. derm. First, Horstadius transplanted Nile blue labeled veg2 cells (the cells that nor mally produce the archenteron) to an ectopic location. This resulted in the autonomous production of a second archenteron by the implanted cells; in addition, perturbations in skeletal morphology, including ectopic skeletal rods, were noted by Horstadius (1935). We have repeated and extended these results by using the chimera technique devised by Wray and McClay (1988) to introduce single rhodamine-labeled veg 2 cells or their descendants into unlabeled host embryos. When such cells are incorporated ectopically, they produce an additional archenteron, but also induce two new bilat eral sites of spicule formation. The ecto derm underneath the new skeletal elements as well as the PMCs that form them are unlabeled (i.e., from the host), and so it appears that lateral induction of host ecto derm by the incorporated cell produces two FIG. 7. Implantation of RITC-labeled macromere descendants results in induction of additional patterning sites for primary mesenchyme. For methods, see the text. (a) Bright field, showing ectopic, bilateral skeletal elements (small arrows), and an ectopic archenteron (large arrow). (b) Epifluorescence demonstrates that only the archenteron cells derive from implanted tissue; the extra skeletal elements and the ectoderm that underlies them are not labeled, indicating that they arise from induction and morphopgenetic movements of host tissue. Bar = 20 !lm. 368 JEFF HARDIN stomodeal invaginatiO~ IA\~Q ~'" ball isolate at mesenchyme blastula stage /l8 lDJectRITC ~.~~ j) isolate at early gastrula stage oral hood and mouth ~ "reconstituted" oral hood oral hood and mouth !lQ FIG. 8. . Summary of experiments demonstrating that the site where oral structures form is fixed by the early gastrula stage. When animal caps are isolated at the mesenchyme blastula stage, they do not produce a stomodeal invagination, and a mouth forms via regulation within the vegetal fragment. In contrast, when the isolation is performed at the early gastrula stage, the animal fragment produces a stomodeal invagination, and can support normal patterning of PMCs when they are transplanted into it. Conversely, the vegetal fragment is no longer capable of forming a mouth. new ectodermal patterning centers (Fig. 7; Hardin, Benink, and Wray, manuscript in preparation). The oral field is restricted to its normal site by the early gastrula stage The final skeletal pattern produced by the sea urchin larva is considerably more com plex than two bilaterally disposed rods. The pattern produced at various locations within the larva is regionally specific as well as spe cies-specific (Ettensohn, 1990; Armstrong and McClay, 1994). How are regional aspects ofskeletal pattern elaborated? Although final conclusions cannot yet be drawn regarding the entire pattern that is produced, several experiments indicate how and when pattern information is laid down in the oral region of the larva. Again, classical experiments by Horstadius hint at cell interactions that are important. He found that the cells destined to give rise to most of the ectoderm (the ani and an 2 tiers in his terminology) formed Dauerblastulae when isolated at the 32- or 64-cell stage. However, when the next more vegetal (veg l ) tier is included, the an/an 2 progeny form a stomodeal invagination (the ectoderm's contribution to the mouth; Hor stadius, 1935). This suggests that interac tions between the veg l progeny and adjacent tiers result in induction of the oral field. More recently, we have performed addi tional isolations to address this question (Hardin and Armstrong, manuscript in preparation). By isolating animal poles at progressively later stages, we have been able to determine the time during which the pat tern information resident in the oral region becomes fixed. When the animal pole is removed prior to the early gastrula stage, the remaining vegetal tissue can regulate to produce a new site of mouth formation, whereas the animal hemisphere does not produce a stomodeum. However, if the experiment is performed somewhat later, at the early gastrula stage, the animal pole frag ment produces a stomodeal invagination, and the remainder of the embryo is unable to produce a mouth. Transplantation of TARGET RECOGNITION BY MESENCHYME DURING SEA URCHIN GASTRULATION PMCs into animal hemispheres isolated at various times indicates that the pattern information required for the production of the parallel skeletal rods flanking the mouth is coordinately regulated with the oral field (these experiments are summarized in Fig. 8). When isolated at progressively later stages, there is a sharp increase in the ability ofanimal pole fragments to support normal patterning (Hardin and Armstrong, in prep aration). CONCLUSIONS The studies we have performed allow us to draw several conclusions regarding how mesenchymal cells find their targets in early embryos. First, in many cases interactions between mesenchyme cells and their targets appear to be highly local. In the case of SMCs, where filopodial dynamics have been examined carefully, these local interactions appear to be contact-mediated, although our experiments do not rule out short-range dif fusible or graded signals (see Hardin and McClay, 1990 for further discussion). Sec ond, target recognition is a specific cell inter action, rather than a general adhesive event. This conclusion is based on the lack of attachment of several other mesenchymal cell types to the animal pole targets for PMCs and SMCs we have identified, and the very different ways in which these two popula tions of cells interact with the ectoderm of the oral field. Third, successive cell inter actions between founder cells and their progeny in the early embryo appear to be responsible for establishing the target regions we have identified, at least in the case of two large clusters of PMCs in the ventro lateral regions of the embryo and mesen chyme in the oral region. The inductive and! or inhibitory events associated with the establishment of such pattern information has not been characterized, although estab lished signal transduction pathways within the cell will undoubtedly be involved (see the article by Wilt, 1995; Cameron et al., 1994). The specific molecular basis for the guidance cues we have identified remains unclear. Putative receptor-ligand interac tions that might mediate specific target rec ognition events await identification at the 369 molecular level. Nevertheless, the cellular interactions we have identified serve as an experimental framework defining processes that candidate guidance and regulatory mol ecules may control. ACKNOWLEDGMENTS This work was supported by a Lucille P. Markey Scholar Award in the Biomedical Sciences, a NSF grant and by a NSF Young Investigator Award to the author. REFERENCES Alberga, A., J. L. Boulay, E. Kempe, C. Dennefeld and M. Haenlin. 1991. The snail gene required for mesoderm formation in Drosophila is expressed dynamically in derivatives ofall three germ layers. Development III :983-992. Armstrong, N. and D. R. McClay. 1994. Skeletal pattern is specified autonomously by the primary mesenchyme cells in sea urchin embryos. Dev. BioI. 162:329-338. Armstrong, N., J. Hardin and D . R. McOay. 1993. Cell-cell interactions regulate skeleton formation in the sea urchin embryo. Development 119:833 840. Benson, S. C. and F. H. Wilt. 1992. Calcification of spicules in the sea urchin embryo. In E. Bonucci (ed.), Calcification in biological systems. pp. 157 178. eRC Press, Boca Raton. Boucaut, J.-c., T. Darrib~re, D. L. Sill, J.-F. Riou, K. E. Johnson and M. Delarue. 1991. Amphibian gastrulation: The molecular bases of mesodermal cell migration in urodele embryos. In R. Keller, W. H. Clark, and F. Griffin (ed.), Gastrulation: Movements. pal/erns. and molecules. pp. 169-184. Plenum Press, New York. Boulay, J. c., C. Dennefield and A. Alberga. 1987. The Drosophila developmental gene snail encodes a protein with nucleic acid binding fingers. Nature 330:395-398. Cameron, R. A. and E. H. Davidson. 1991. Cell type specification during sea urchin development. Trends Gen. 7:212-218. Cameron, R. A., S. E. Fraser, R. J. Britten and E. H. Davidson. 1991 . Macromere cell fates during sea urchin development. Development 113:1085 1091. Cameron, R. A., B. R. Hough-Evans, R. J. Britten and E. H. Davidson. 1987. Lineage and fate of each blastomere of the eight-cell sea urchin embryo. Genes Dev. 1:75-84. Cameron, R. A., L. C. Smith, R. J. Britten and E. H. Davidson. 1994. Ligand-dependent stimulation of introduced mammalian brain receptors alters spicule symmetry and other morphogenetic events in sea urchin embryos. Mech. of Dev. 45:31-47 . Dan, K. and D. Inaba. 1968. Echinoderma. In M. Kume, and K. Dan (ed.), Invertebrate embryology. pp. 280-332. NOLIT Publishing House, Belgrade. Davidson, E. H. 1989. Lineage-specific gene expres 370 JEFF HARDIN sion and the regulative capacities of the sea urchin embryo: A proposed mechanism . Development 105:421-445 . Davidson, E. H. 1993. Later embryogenesis: Regu latory circuitry in morphogenetic fields. Devel opment 118:665-690. Decker, G . L. and W. J . Lennarz. 1988. Skeletoge nesis in the sea urchin embryo. Develop. 103:231 247. Driesch, H. 1900. Die isolirten Blastomeren des Echinidenkeimes. Arch. Entwicklungsmech. Org. 10:361-410. Ettensohn, C. A. 1990. The regulation of primary mesenchyme cell patterning. Dev. BioI. 140:261 271. Ettensohn, C. A. 1991 . Primary mesenchyme cell migration in the sea urchin embryo. In R. Keller, W. H. Clark Jr. , and F. Griffin (ed.), Gastrulation: Movements, patterns and molecules, pp. 289-304 . Plenum, New York. Ettensohn, C. A. and E. P. Ingersoll. IQ92. Morpho genesis of the sea urchin embryo. In E. F. Ros somando and S. Alexander (ed.), Morphogenesis: An analysis of the development ofbiological form, pp. 189-262. Marcel, Dekker, New York. Ettensohn, C. A. and D. R. McClay. 1986. The reg ulation of primary mesenchyme cell migration in the sea urchin embryo: Transplantations of cells and latex beads. Dev. BioI. 117:380-391. Gustafson, T. 1963. Cellular mechanisms in the mor phogenesis of the sea urchin embryo. Cell contacts within the ectoderm and between mesenchyme and ectoderm cells. Exp. Cell Res. 32:570-589 . Gustafson, T . 1969. Cell recognition and cell contacts during sea urchin development. In R. T . Smith, and R. A. Good (ed.), Cellular recognition, pp. 47 60. Appleton-Century-Crofts, New York. Gustafson, T. and H. Kinnander. 1960. Cellular mechanisms in morphogenesis of the sea urchin gastrula. The oral contact. Exp. Cell Res. 21:361 373. Gustafson, T . and L. Wolpert. 1963. The cellular basis of morphogenesis and sea urchin develop ment. Int. Rev. Cyt. 15: 139-214. Gustafson, T . and L. Wolpert. 1967. Cellular move ment and contact in sea urchin morphogenesis. BioI. Rev. 42:442-498. Hammerschmidt, M. and C. Niisslein- Volhard. 1993. The expression of a zebrafish gene homologous to Drosophila snail suggests a conserved function in invertebrate and vertebrate gastrulation. Devel opment 119:1107-1118. Hardin, J. 1988. The role of secondary mesenchyme cells during sea urchin gastrulation studied by laser ablation. Development 103:317- 324. Hardin, J. 1989. Local shifts in position and polar ized motility drive cell rearrangement during sea urchin gastrulation. Dev. BioI. 136:430-445 . Hardin, 1. 1990. Context-sensitive cell behaviors during gastrulation. Sem. Dev. BioI. 1:335-345. Hardin, J. 1994a. Local cell interactions and the con trol ofgastrulation in the sea urchin embryo. Sem. Dev. BioI. 5:77-84. Hardin, J. I 994b. The cellular basis of sea urchin gastrulation. Curro Topics. Dev. BioI. (In press) Hardin, J. and D. R. McClay. 1990. Target recog nition by the archenteron during sea urgin gastru lation. Dev. BioI. 142:87-105 . Hardin, J.,J. A. Coffman,S. D. BlackandD. R. McClay. 1992. Commitment along the dorsoventral axis of the sea urchin embryo is altered in response to NiCI,. Development 116:671-685. Ho, R. K. 1992. Cell movements and cell fate during zebrafish gastrulation. Development 1992 Sup plement:65-73 . Horstadius, S. 1935. Uberdie Determination im Ver laufe der Eiachse bei Seeigeln. Pubbl. Staz. Zool. Napoli 14:251-429. Horstadius, S. 1939. The mechanics of sea urchin development, studied by operative methods. Bio. Rev. Cambridge Phil. Soc. 14: 132-179. Horstadius, S. 1973. Ex perimental embryology of echinoderms. Clarendon Press, Oxford. Ip, Y. T. , R. E. Park, D. Kosman, K. Yazdanbakhsh and M. Levine. 1992. dorsal-twist interactions establish snail expression in the presumptive mesoderm of the Drosophila embryo. Genes & Development 6:1518-1530. Keller, R . and R. WinkIbauer. 1992. Cellular basis ofamphibian gastrulation. Curr. Topics Dev. BioI. 27:39-89. Khaner, O. and F. Wilt. 1990. The influence of cell interactions and tissue mass on differentiation of sea urchin mesomeres. Development 109:625-634. Leptin, M. 1991. twist and snail as positive and neg ative regulators during Drosophila mesoderm development. Genes & Development 5: 1568-1576. Livingston, B. T. and F. H. Wilt. 1990. Determina tion of cell fate in sea urchin embryos. Bioessays 12:115-119. McClay, D. R., N . A. Armstrong and J. Hardin. 1992. Pattern formation during gastrulation in the sea urchin embryo. 1992 Supplement:33-41 . Nieto, A. M., M. G. Sargent, D. G . Wilkinson and J. Cooke. 1994. Control of cell behavior during vertebrate development by Slug, a zinc finger gene. Science 264:835-839. Nieto, M. A. , M. F. Bennett, M. G. Sargent and D. G. Wilkinson. 1992. Cloning and developmental expression ofSna, a murine homologue ofthe Dro sophila snail gene. Development 116:227-237. Okazaki, K. , T . Fukushi and K. Dan . 1962. Cyto embryological studies of sea urchins. IV . Corre lation between the shape of the ectodermal cells and the arrangement of the primary mesenchyme cells in sea urchin larvae. Acta Embryol. Morphol. Exp. 5: 17-31. Ransick , A. and E. H . Davidson. 1993. A complete second gut induced by transplanted micromeres in the sea urchin embryo. Science 259: 1134-1138. Sargent, M. G. and M. F. Bennett. 1990. Identifica tion in Xenopus of a structural homologue of the Drosophila gene snail. Development 109:967-973. Smith, D. E., F. Franco del Amo and T . Gridley. 1992. Isolation ofSna, a mouse gene homologous to the Drosophila genes snail and escargot: Its expression pattern suggests multiple ro.les during postimplan tation development. Development 116: 1033-1039. Solursh, M. 1986. Migration of sea urchin primary mesenchyme cells. In L. Browder (ed.), The cel TARGET RECOGNITION BY MESENCHYME DURING lular basis ofmorphogenesis, pp. 391 ~31. Plenum Press, New York. Stern, C D. 1990. The marginal zone and its con tribution to the hypoblast and primitive streak of the chick embryo. Development 109:667-682. Stern, C. D. 1991. Mesodenn fonnation in the chick embryo, revisited. In R. Keller, W. H. Clark, and F. Griffin (ed.), Gastrulation: Movements pallerns, and molecules, pp. 29~ I. Plenum Press, New York. Takahashi, M. M. and K. Okazaki. 1979. Total cell number and number of the primary mesenchyme cells in w'lole, '12, and '/4 larvae of Clypleaster japonicus. Dev. Growth Differ. 21 :553-566. Thisse, C, B. Thisse, T. F. Schilling and J. H. Pos tlethwait. 1993. Structure of the zebrafish snaill gene and its expression in wild-type, spadetail and no tail mutant embryos. Development 119: 1203 1215. Trinkaus, J. P. 1984. Cell into organs: The forces that shape the embryo. 2nd ed. Prentice-Hall, Engle wood Cliffs. Trinkaus, J. P., M. Trinkaus and R. D. Fink. 1991. In vivo analysis of convergent cell movements in the genn ring of Fundulus. In R. Keller, W. H. Clark, and F. Griffin (ed.), Gastrulation: Move ments, pallerns, and molecules, pp. 121-134. Ple num Press, New York. Whiteley, M., P. D. Noguchi, S. M. Sensabaugh, W. SEA URCHIN GASTRULATION 371 F. Odenwald and 1. A. Kassis. 1992. The Dro sophila gene escargot encodes a zinc finger motif found in snail-related genes. Mechanisms of Development 36:117-127. Wilt, F. H. 1987. Detennination and morphogenesis in the sea urchin embryo. Develop. 100:559-575. Wilt, F. H., B. N. Livingston and O. Khaner. 1995. Cell interactions during early sea within devel opment Amer. Zoo!. 35:353-357. Winklbauer, R., A. Selchow, M. Nagel, C Stoltz and B. Angres. 1991. Mesodenn cell migration in the Xenopus gastrula. In R. KeUer, W. H. Clark, and F. Griffin (ed.), Gastrulation: Movements, pallerns, and molecules, pp. 147-168. Plenum Press, New York. Wolpert, L. and T. Gustafson. 1961. Studies on the cellular basis of morphogenesis in the sea urchin embryo. Gastrulation in vegetalized larvae. Exp. Cell Res. 22:437~49. Wounns, J. P. 1972. The developmental biology of annual fishes . II. Naturally occurring dispersion and reaggregation of blast orne res during the devel opment ofannual fish eggs. J. Exp. Zoo!. 182: 169 200. Wray, G. A. and D. R. McClay. 1988. The origin of spicule-fonning cells in a "primitive" sea urchin (Eucidaris tribuloides) which appears to lack pri mary mesenchyme. Development 103:305-315. .-.